Compare commits

..

821 commits

Author SHA1 Message Date
3661a597fa Merge pull request 'feat: add use_local_tz configuration' (#908) from ragazenta/garage:feat/local-timezone into main
Reviewed-on: Deuxfleurs/garage#908
Reviewed-by: maximilien <me@mricher.fr>
2024-12-01 13:23:24 +00:00
0fd3c0e794
doc: add use_local_tz configuration 2024-11-25 10:35:00 +07:00
4c1bf42192
feat: add use_local_tz configuration
Used in lifecycle_worker to determine midnight time
2024-11-23 05:51:12 +07:00
906c8708fd Merge pull request 'add extraVolumes and extraVolumeMounts to helm chart' (#896) from eugene-davis/garage:main into main
Reviewed-on: Deuxfleurs/garage#896
Reviewed-by: maximilien <me@mricher.fr>
2024-11-19 22:23:13 +00:00
747889a096 Merge pull request 'Update Python SDK documentation' (#887) from cryptolukas/garage:fix-python-sdk-docs into main
Reviewed-on: Deuxfleurs/garage#887
2024-11-19 09:15:03 +00:00
feb09a4bc6 Merge pull request 'doc: update mastodon media header pruning section' (#888) from teutat3s/garage:doc-update-mastodon-media into main
Reviewed-on: Deuxfleurs/garage#888
2024-11-19 09:14:34 +00:00
aa8bc6aa88 Merge pull request 'doc: add Triplebit's use-case' (#901) from jonah/garage:triplebit into main
Reviewed-on: Deuxfleurs/garage#901
Reviewed-by: maximilien <me@mricher.fr>
2024-11-17 13:43:49 +00:00
aba7902995
doc: add Triplebit's use-case 2024-11-15 16:27:46 -06:00
78de7b5bde Merge pull request 'fix bit/byte inversion in rpc secret error message' (#898) from trinity-1686a/garage:rpc-comment into main
Reviewed-on: Deuxfleurs/garage#898
2024-11-07 11:11:12 +00:00
9bd9e392ba fix bit/byte inversion in rpc secret error message 2024-11-07 00:29:26 +01:00
116ad479a8
add extraVolumes and extraVolumeMounts to helm chart 2024-10-26 21:14:08 +02:00
teutat3s
b6a58c5c16
doc: update mastodon media header pruning section
This is now possible since the upstream issue has been resolved.
https://github.com/mastodon/mastodon/issues/9567
2024-10-17 20:59:21 +02:00
2b0bfa9b18 the old value do not work out of the box 2024-10-14 17:20:26 +02:00
a18b3f0d1f Merge pull request 'Garage v1.0.1' (#881) from rel-v1.0.1 into main
Reviewed-on: Deuxfleurs/garage#881
2024-09-22 13:02:02 +00:00
7a143f46fc
Bump to version 1.0.1 2024-09-22 14:25:32 +02:00
c731f0291a Merge pull request 'fix logic in garage layout skip-dead-nodes + fix typo (fix #879)' (#880) from fix-skip-dead-nodes into main
Reviewed-on: Deuxfleurs/garage#880
2024-09-22 12:01:49 +00:00
34453bc9c2
fix logic in garage layout skip-dead-nodes + fix typo (fix #879) 2024-09-22 13:47:27 +02:00
6da1353541 Merge pull request 'Don't fetch old values in cross-partition transactional inserts' (#877) from withings/garage:perf/kv/insert-no-return-cross-partition into main
Reviewed-on: Deuxfleurs/garage#877
2024-09-14 15:57:27 +00:00
bd71728874
Tests: don't expect old value after transactional insert 2024-09-12 10:50:53 +02:00
51ced60366
Don't fetch old values in cross-partition transactional inserts 2024-09-12 10:26:28 +02:00
586957b4b7 Merge pull request 'KV: don't retrieve values for write ops' (#873) from marvinj97/garage:perf/kv/insert-no-return into main
Reviewed-on: Deuxfleurs/garage#873
Reviewed-by: Alex <alex@adnab.me>
2024-09-10 09:06:29 +00:00
8d2bb4afeb Merge pull request 'Typo' (#875) from faust/garage:doc2 into main
Reviewed-on: Deuxfleurs/garage#875
2024-09-10 09:03:31 +00:00
c26f32b769
Typo
And remove trailing white space.
2024-09-10 09:34:59 +02:00
8062ec7b4b test: fix db tests 2024-09-04 19:24:36 +02:00
eb416a02fb dont assert deletion count in sqlite KV adapter 2024-09-04 18:51:51 +02:00
74363c9060 perf(kv): dont retrieve values for write ops
see Deuxfleurs/garage#851
2024-09-04 18:45:17 +02:00
615698df7d Merge pull request 'update compiler to Rust 1.77' (#866) from rust-1.77 into main
Reviewed-on: Deuxfleurs/garage#866
2024-08-26 19:08:00 +00:00
7061fa5a56
use rust 1.77 in nix/compile.nix 2024-08-26 19:19:16 +02:00
8881930cdd
update nixpkgs and rust-overlay sources in flake.nix 2024-08-26 19:19:16 +02:00
52f6c0760b Merge pull request 'update crate time (fix #849)' (#865) from update-time into main
Reviewed-on: Deuxfleurs/garage#865
2024-08-26 16:20:04 +00:00
5b0602c7e9
update crate time (fix #849) 2024-08-26 18:11:21 +02:00
182b2af7e5 Merge pull request 'api servers: kill opened connections after SIGINT after 10s deadline (fix #806)' (#864) from exit-deadline into main
Reviewed-on: Deuxfleurs/garage#864
2024-08-25 18:34:55 +00:00
baf32c9575
api servers: kill opened connections after SIGINT after 10s deadline (fix #806) 2024-08-25 20:04:56 +02:00
3dda1ee4f6 Merge pull request 'fix build when lmdb feature is disabled (fix #800)' (#863) from fix-800 into main
Reviewed-on: Deuxfleurs/garage#863
2024-08-25 10:00:40 +00:00
aa7ce9e97c
fix build when lmdb feature is disabled (fix #800) 2024-08-25 11:42:37 +02:00
8d62616ec0 Merge pull request 'layout: discard old info when it is completely out-of-date (fix #841)' (#861) from fix-841 into main
Reviewed-on: Deuxfleurs/garage#861
2024-08-24 11:12:39 +00:00
bd6fe72c06 Merge pull request 'Quick start: mention Docker (replace #803)' (#862) from dougreeder into main
Reviewed-on: Deuxfleurs/garage#862
2024-08-24 11:07:46 +00:00
4c9e8ef625
doc: clarify quick start on using docker 2024-08-24 13:07:02 +02:00
3e711bc110 Merge pull request 'don't modify postobject request before validating policy' (#850) from trinity-1686a/garage:fix-acl-postobject into main
Reviewed-on: Deuxfleurs/garage#850
2024-08-24 10:49:14 +00:00
7fb66b4944
layout: discard old info when it is completely out-of-date (fix #841) 2024-08-24 12:38:56 +02:00
679ae8bcbb Merge pull request 'Set "no read ahead" on LMDB to improve performances when "LMDB size" > "RAM size"' (#855) from fix-lmdb-no-read-ahead into main
Reviewed-on: Deuxfleurs/garage#855
Reviewed-by: Alex <alex@adnab.me>
2024-08-18 12:25:35 +00:00
2a93ad0c84
force flag "no read ahead" on LMDB 2024-08-17 21:17:15 +02:00
f190032589 don't modify postobject request before validating policy 2024-08-10 20:10:47 +02:00
3a87bd1370 Merge pull request 'Improve error message for malformed RPC secret key' (#846) from improve-secret-error-message into main
Reviewed-on: Deuxfleurs/garage#846
Reviewed-by: Quentin <quentin@dufour.io>
2024-08-09 06:47:11 +00:00
9302cd42f0 Improve error message for malformed RPC secret key 2024-08-08 23:05:24 +00:00
060ad0da32 docs: Update LMDB website 2024-08-06 21:47:14 +00:00
a5ed1161c6 Merge pull request 'Add environment variable dict to helm chart.' (#843) from Benjamin/garage:main into main
Reviewed-on: Deuxfleurs/garage#843
Reviewed-by: maximilien <me@mricher.fr>
2024-08-06 21:45:35 +00:00
Benjamin von Mossner
222674432b This commit adds an environment dict to garage helm chart. Using it, env variables can be set into the garage container environment, useful to set eg. GARAGE_ADMIN_TOKEN or GARAGE_METRICS_TOKEN 2024-07-25 11:42:13 +02:00
070a8ad110 Merge pull request 'doc: fix typo' (#831) from Armael/garage:typo into main
Reviewed-on: Deuxfleurs/garage#831
2024-06-18 12:40:32 +00:00
770384cae1 Merge pull request 'add rpc_public_addr_subnet config option' (#817) from flokli/garage:rpc_public_addr_subnet into main
Reviewed-on: Deuxfleurs/garage#817
Reviewed-by: Alex <alex@adnab.me>
2024-06-18 12:40:07 +00:00
a0f6bc5b7f add rpc_public_addr_subnet config option
In case `rpc_public_addr` is not set, but autodiscovery is used, this
allows filtering the list of automatically discovered IPs to a specific
subnet.

For example, if nodes should pick *their* IP inside a specific subnet,
but you don't want to explicitly write the IP down (as it's dynamic, or
you want to share configs across nodes), you can use this option.
2024-06-05 08:41:36 +02:00
Armaël Guéneau
88c734bbd9 typo 2024-06-04 15:34:02 +02:00
d38509ef4b Merge pull request 'adding the ability to change the default podManagementPolicy for StatefulSets' (#823) from bodaciousbiscuits/garage:main into main
Reviewed-on: Deuxfleurs/garage#823
Reviewed-by: maximilien <me@mricher.fr>
2024-05-25 18:35:53 +00:00
bodaciousbiscuits
39b37833c5 adding the ability to change the default podManagementPolicy for StatefulSets 2024-05-19 21:31:19 -05:00
a2c1de646b Merge pull request 'cli: clarify garage block is node-local' (#813) from flokli/garage:block-node-local into main
Reviewed-on: Deuxfleurs/garage#813
2024-05-12 08:53:26 +00:00
15847a636a cli: clarify garage block is node-local
Prevents some of the confusion from
Deuxfleurs/garage#810.
2024-05-07 07:42:33 +00:00
123d3e1f04 Merge pull request 'flake.nix: add rust-analyzer to devShells.full, expose devShells.full in shell.nix' (#816) from flokli/garage:shell-fixes into main
Reviewed-on: Deuxfleurs/garage#816
2024-04-23 18:54:26 +00:00
a6e4b96ca9 shell.nix: expose devShellFull
This allows accessing devShells.full from shell.nix.
2024-04-23 11:59:37 +03:00
b442b0e35e devShells.full: add rust-analyzer 2024-04-23 11:59:37 +03:00
0c3b198b22 Improves Quick Start for users not using Linux 2024-04-10 16:42:10 -04:00
33c2086d9e Merge pull request '[fix-doc] fix broken references in documentation' (#802) from fix-doc into main
Reviewed-on: Deuxfleurs/garage#802
2024-04-10 15:49:03 +00:00
5ad1e55ccf [fix-doc] fix broken references in documentation 2024-04-10 17:47:34 +02:00
1779fd40c0 Merge pull request 'Garage v1.0' (#683) from next-0.10 into main
Reviewed-on: Deuxfleurs/garage#683
2024-04-10 15:23:12 +00:00
ff093ddbb8
Merge branch 'main' into next-0.10 2024-04-10 14:38:14 +02:00
90e3c2af91
[next-0.10] small updates to mention Garage v0.9.4 2024-04-10 14:35:30 +02:00
b47706809c Merge pull request 'fix typo in doc' (#799) from fix-typo into main
Reviewed-on: Deuxfleurs/garage#799
2024-04-08 15:09:27 +00:00
126e0f47a3 fix typo in doc 2024-04-08 17:08:44 +02:00
738bb2f09c Merge pull request 'Garage v0.9.4' (#798) from rel-0.9.4 into main
Reviewed-on: Deuxfleurs/garage#798
2024-04-08 09:57:03 +00:00
7dd7cb5759
[rel-0.9.4] upgrade version to v0.9.4 2024-04-08 11:18:19 +02:00
8b663d8c5b Merge pull request 'jepsen testing of Garage v1.0.0-rc1' (#796) from jepsen-1.0rc1 into main
Reviewed-on: Deuxfleurs/garage#796
2024-04-05 21:05:58 +00:00
c051db8204 jepsen testing of Garage v1.0.0-rc1 2024-04-05 22:57:00 +02:00
50669b3e76
[next-0.10] bump helm chart version 2024-04-03 14:19:59 +02:00
e5838b4837 Merge pull request '[doc-units] document how interval value is parsed' (#795) from doc-units into main
Reviewed-on: Deuxfleurs/garage#795
2024-04-03 12:15:18 +00:00
87dfaf2eb9
[doc-units] document how interval value is parsed 2024-04-03 14:14:13 +02:00
554437254e
[next-0.10] Add migration guide for v1.0 2024-03-28 18:45:06 +01:00
afad62939e
[next-0.10] bump version number to 1.0 2024-03-28 15:19:44 +01:00
8bfc16ba7d
Merge branch 'main' into next-0.10 2024-03-28 15:01:05 +01:00
ecf641d88c Merge pull request 'Fix unbounded buffering when one node has slower network' (#792) from fix-buffering into main
Reviewed-on: Deuxfleurs/garage#792
2024-03-28 12:40:27 +00:00
75cd14926d Merge pull request 'CI: properly cleanup between garage integration tests' (#793) from fix-ci into main
Reviewed-on: Deuxfleurs/garage#793
2024-03-28 12:37:18 +00:00
e1dc84e123
[fix-ci] CI: properly cleanup between garage integration tests 2024-03-28 13:08:42 +01:00
85f580cbde
[fix-buffering] change request sending strategy and fix priorities
remove LAS, priorize new requests but otherwise just do standard queuing
2024-03-27 16:22:40 +01:00
0d3e285d13
[fix-buffering] implement block_ram_buffer_max to avoid excessive RAM usage 2024-03-27 16:22:40 +01:00
25c196f34d
[next-0.10] admin api: fix logic in get cluster status 2024-03-27 13:55:49 +01:00
4eba32f29f
[next-0.10] layout helper: rename & clarify updates to update trackers 2024-03-27 13:47:06 +01:00
32f1786f9f
[next-0.10] cache layout check result 2024-03-27 13:37:20 +01:00
01a0bd5410
[next-0.10] remove impl Deref for LayoutHelper 2024-03-27 13:32:13 +01:00
c0eeb0b0f3
[next-0.10] fixes to k2v rpc + comment fixes 2024-03-27 10:44:03 +01:00
51d11b4b26
[next-0.10] doc: 2 changes
- rewrite section on encryption to mention SSE-C
- change to real-world to make it closer to main branch
2024-03-27 10:10:45 +01:00
f7cd4eb600
Merge branch 'main' into next-0.10 2024-03-26 16:34:40 +01:00
95eb8808e8 Merge pull request 'Disable more K2V tests' (#791) from disable-k2v-test into main
Reviewed-on: Deuxfleurs/garage#791
2024-03-26 15:33:46 +00:00
e0a4fc097a
[disable-k2v-test] remove obsolete k2v test script 2024-03-26 16:24:44 +01:00
73551e9a2d
[disable-k2v-test] disable the other k2v poll test 2024-03-26 16:24:26 +01:00
80f81fa6f3 Merge pull request '[disable-k2v-test] disable tests::k2v::test_poll_item as it is not 100% reliable' (#789) from disable-k2v-test into main
Reviewed-on: Deuxfleurs/garage#789
2024-03-26 14:57:46 +00:00
f267609343
[disable-k2v-test] disable tests::k2v::test_poll_item as it is not 100% reliable 2024-03-26 15:39:17 +01:00
cdde0f19ee Merge pull request 'checksum algorithms' (#787) from s3-checksum into next-0.10
Reviewed-on: Deuxfleurs/garage#787
2024-03-26 14:24:58 +00:00
74949c69cb
[s3-checksum] implement x-amz-checksum-* headers 2024-03-26 15:01:34 +01:00
7e0107c47d Merge pull request 'Fixes to garage_net peering manager' (#786) from net-fixes into next-0.10
Reviewed-on: Deuxfleurs/garage#786
2024-03-21 10:26:36 +00:00
3844110cd0
[net-fixes] netapp peer exchange: send only currently connected address 2024-03-21 10:50:44 +01:00
961b4f9af3
[net-fixes] fix issues with local peer address (fix #761) 2024-03-21 10:45:34 +01:00
5225a81dee
[net-fixes] peering: only count node IDs and not addresses in hash 2024-03-21 09:47:04 +01:00
e835196940 Merge pull request 'Add marker files in data directories (fix #601)' (#785) from check-data-dir into main
Reviewed-on: Deuxfleurs/garage#785
2024-03-20 16:53:47 +00:00
ba33bb31f1
[check-data-dir] add marker files in data directories (fix #601) 2024-03-20 15:20:25 +01:00
30abf7e086 Merge pull request 'Add support to logging to syslog (based on patch by @jirutka)' (#784) from syslog into main
Reviewed-on: Deuxfleurs/garage#784
2024-03-20 13:54:18 +00:00
84018be862
[syslog] warning when syslog support is not enabled 2024-03-20 14:39:04 +01:00
091e693670
[syslog] document environment variables 2024-03-20 14:39:04 +01:00
fe8a7819fa
[syslog] Add support to logging to syslog
Original patch by Jakub Jirutka for Alpine Linux port.
2024-03-20 14:22:18 +01:00
ce69dc302c
Merge branch 'main' into next-0.10 2024-03-19 17:17:46 +01:00
26310f3242 Merge pull request 'CLI: allow manipulating buckets by prefixes of their full IDs' (#783) from bucket-id-prefix into main
Reviewed-on: Deuxfleurs/garage#783
2024-03-19 16:17:16 +00:00
65853a4863 Merge pull request 'block refcount repair' (#782) from block-ref-repair into next-0.10
Reviewed-on: Deuxfleurs/garage#782
2024-03-19 15:59:19 +00:00
783b586de9
[bucket-id-prefix] CLI: allow manipulating buckets by prefixes of their full IDs 2024-03-19 16:57:51 +01:00
3eab639c14
[block-ref-repair] mention garage block repair-rc in documentation 2024-03-19 16:24:34 +01:00
3165ab926c
[block-ref-repair] rename rc's rc field to rc_table 2024-03-19 16:20:22 +01:00
dc0b78cdb8
[block-ref-repair] Block refcount recalculation and repair
- We always recalculate the reference count of a block before deleting
  it locally, to make sure that it is indeed zero.

- If we had to fetch a remote block but we were not able to get it,
  check that refcount is indeed > 0.

- Repair procedure that checks everything
2024-03-19 16:20:22 +01:00
693b89b94b Merge pull request 'Update WinSCP link in documentation' (#781) from stefano/garage:main into main
Reviewed-on: Deuxfleurs/garage#781
2024-03-19 09:34:16 +00:00
cf344d73d5 Update WinSCP link in documentation
Update link to new wiki location. See Deuxfleurs/garage#780
2024-03-19 09:21:50 +00:00
0038ca8a78
Merge branch 'main' into next-0.10 2024-03-18 20:19:30 +01:00
1a0bffae34 Merge pull request 'Use connection pooling in sqlite backend' (#779) from sqlite-r2d2 into main
Reviewed-on: Deuxfleurs/garage#779
2024-03-18 19:02:36 +00:00
b55f52a9b7
[sqlite-r2d2] run integration test with all db engines 2024-03-18 18:31:35 +01:00
e8f9718ccd
[sqlite-r2d2] implement connection pooling in sqlite backend 2024-03-18 18:05:25 +01:00
fd2e19bf1b Merge pull request 'metadata db snapshotting' (#775) from db-snapshot into main
Reviewed-on: Deuxfleurs/garage#775
2024-03-15 13:17:53 +00:00
8cf3d24875
[db-snapshot] documentation for metadata db snapshots 2024-03-15 13:51:31 +01:00
a68c37555d
[db-snapshot] add garage meta snapshot cli operation 2024-03-15 13:51:31 +01:00
1e42808a59
[db-snapshot] implement meta_auto_snapshot_interval 2024-03-15 13:51:31 +01:00
8dff278b72
[db-snapshot] Implement db snapshotting logic in garage_db 2024-03-15 10:57:22 +01:00
a80ce6ab5a Merge pull request 'disable_scrub configuration option' (#774) from disable-scrub into main
Reviewed-on: Deuxfleurs/garage#774
2024-03-15 09:22:33 +00:00
990205dc3b
[disable-scrub] document disable_scrub config option 2024-03-14 17:01:16 +01:00
7c86ff6c37
[disable-scrub] implement a disable_scrub configuration option 2024-03-14 17:01:16 +01:00
62b01d8705 Merge pull request 'Doc: be slightly more critical of LMDB' (#773) from doc-updates into main
Reviewed-on: Deuxfleurs/garage#773
2024-03-14 15:43:30 +00:00
422d45b659
[doc-updates] from source: fix default feature list 2024-03-14 16:35:15 +01:00
a7dddebedd
[doc-updates] doc: be slightly more critical of LMDB 2024-03-14 16:31:22 +01:00
81191d2d92 Merge pull request 'Remove Sled' (#767) from rm-sled into next-0.10
Reviewed-on: Deuxfleurs/garage#767
2024-03-12 10:45:57 +00:00
2795b53b8b
[rm-sled] factorize some code in sqlite backend 2024-03-12 11:15:26 +01:00
32aa246300
[rm-sled] Make proper use of pinning in LMDB adapter + comment unsafe 2024-03-08 17:39:17 +01:00
b942949940
[rm-sled] Implement iterators in sqlite & lmdb transactions
with way too much unsafe code
2024-03-08 16:38:01 +01:00
66c23890c1
[rm-sled] Implement some missing functionality in garage_db 2024-03-08 16:02:58 +01:00
05c92204ec
[rm-sled] Remove counted_tree_hack 2024-03-08 15:09:57 +01:00
2128b5febd Merge pull request 'Remove migration path from Garage v0.5' (#766) from rm-migration into next-0.10
Reviewed-on: Deuxfleurs/garage#766
2024-03-08 13:43:42 +00:00
44454aac01
[rm-sled] Remove the Sled database engine 2024-03-08 14:11:02 +01:00
1ace34adbb
Merge branch 'main' into next-0.10 2024-03-08 13:57:10 +01:00
238545e564 Merge pull request 'Refactoring of db engines' (#765) from factor-db-open into main
Reviewed-on: Deuxfleurs/garage#765
2024-03-08 12:56:40 +00:00
f537f76681
[rm-migration] Remove migration path from Garage v0.5 2024-03-08 13:24:47 +01:00
ec34728b27
[factor-db-open] Combine logic for opening db engines 2024-03-08 12:58:17 +01:00
20c0b4ffb2 Merge pull request 'ReplicationMode -> ConsistencyMode+ReplicationFactor' (#750) from yuka/garage:split-consistency-mode into next-0.10
Reviewed-on: Deuxfleurs/garage#750
2024-03-07 16:32:52 +00:00
2fd13c7d13 Merge pull request 'SSE-C encryption' (#730) from sse-c into next-0.10
Reviewed-on: Deuxfleurs/garage#730
2024-03-07 15:21:37 +00:00
3fcb54e3cf
[sse-c] Remove special case for Content-Type header 2024-03-07 15:43:48 +01:00
e3333f2ac5
[sse-c] Documentation for SSE-C 2024-03-07 15:43:48 +01:00
fa4878bad6
[sse-c] Testing for SSE-C encryption 2024-03-07 15:43:48 +01:00
57acc60082
[sse-c] Implement SSE-C encryption 2024-03-07 15:43:47 +01:00
fe2dc5d51c
Merge branch 'main' into next-0.10 2024-03-07 14:00:34 +01:00
afee8c2207 Merge pull request 'allow utf-8 in headers + add test for object metadata' (#763) from unicode-headers into main
Reviewed-on: Deuxfleurs/garage#763
2024-03-07 12:54:07 +00:00
eab2b81be2
[unicode-headers] allow utf-8 in headers + add test for object metadata 2024-03-07 13:42:01 +01:00
Yureka
c1769bbe69 ReplicationMode -> ConsistencyMode+ReplicationFactor 2024-03-07 12:45:33 +01:00
Yureka
8f86af52ed adjust docs for replication factor 2024-03-05 22:57:08 +01:00
603604cdfc Merge pull request 'refactor: remove max_write_errors and max_faults' (#760) from yuka/garage:remove-max-write-errors into next-0.10
Reviewed-on: Deuxfleurs/garage#760
2024-03-05 21:56:17 +00:00
Yureka
6760895926 refactor: remove max_write_errors and max_faults 2024-03-04 18:39:56 +01:00
bbde9bc912
Merge branch 'main' into next-0.10 2024-03-04 15:56:10 +01:00
3168bb34a0 Merge pull request 'add request context helper' (#751) from yuka/garage:req-ctx into main
Reviewed-on: Deuxfleurs/garage#751
2024-03-04 14:51:05 +00:00
512933a036 Merge pull request 'Garage v0.9.3' (#757) from rel-0.9.3 into main
Reviewed-on: Deuxfleurs/garage#757
2024-03-04 13:26:47 +00:00
8670140358
[rel-0.9.3] Bump version to 0.9.3 2024-03-04 14:00:55 +01:00
5bb69a1257 Merge pull request 'Add API test + fix presigned requests' (#756) from test-presigned into main
Reviewed-on: Deuxfleurs/garage#756
2024-03-04 12:56:02 +00:00
c8e416aaa5
[test-presigned] Use a HeaderMap type for QueryMap 2024-03-04 13:33:14 +01:00
Yureka
fb55682c66 add request context helper 2024-03-04 13:26:39 +01:00
c94bf45cba
Store original-cased query keys alongside query values 2024-03-04 13:03:27 +01:00
7c4f3473af
Lowercase query parameter keys when parsing 2024-03-04 13:03:16 +01:00
b6a91e549b
[test-presigned] Add API test for presigned requests 2024-03-04 13:02:07 +01:00
32d6b4def8 Merge pull request 'Add talk on 2024-02-09 at capitoul.org' (#755) from talk-capitoul into main
Reviewed-on: Deuxfleurs/garage#755
2024-03-04 11:08:23 +00:00
c4de471de1 Merge pull request 'doc: fix typo in connect/backup.md' (#749) from Armael/garage:doc-typo into main
Reviewed-on: Deuxfleurs/garage#749
2024-03-03 13:51:38 +00:00
Armaël Guéneau
16e17375c5 doc: fix typo in connect/backup.md 2024-03-03 13:02:56 +01:00
95ab36aae7 Merge pull request 'Bump version to v0.9.2' (#747) from rel-0.9.2 into main
Reviewed-on: Deuxfleurs/garage#747
2024-03-01 16:20:28 +00:00
6a7623e90d
[rel-0.9.2] Bump version to v0.9.2 2024-03-01 16:54:39 +01:00
70b9904e91 Merge pull request 'AWS signatures v4: don't actually check Content-Type is signed' (#745) from fix-signed-headers into main
Reviewed-on: Deuxfleurs/garage#745
2024-03-01 12:50:15 +00:00
a36248a169
[fix-signed-headers] aws signatures v4: don't actually check Content-Type is signed
This page of the AWS docs indicate that Content-Type should be part of
the CanonicalHeaders (and therefore SignedHeaders) strings in signature
calculation:

https://docs.aws.amazon.com/AmazonS3/latest/API/sig-v4-header-based-auth.html

However, testing with Minio Client revealed that it did not sign the
Content-Type header, and therefore we broke CI by expecting it to be
signed. With this commit, we don't mandate Content-Type to be signed
anymore, for better compatibility with the ecosystem. Testing against
the official behavior of S3 on AWS has not been done.
2024-03-01 13:12:18 +01:00
b8c7a560ef Merge pull request 'Fix potential timing side-channels in authentication mechanisms' (#737) from fix-auth-ct-eq into main
Reviewed-on: Deuxfleurs/garage#737
2024-02-29 14:04:38 +00:00
d3cf560e5c Merge pull request 'Docs: add default metrics_token in quick start + uniformize use of base64' (#739) from doc-default-token into main
Reviewed-on: Deuxfleurs/garage#739
2024-02-29 12:20:24 +00:00
73b11eb17c
[doc-default-token] add default metrics_token in quick start + uniformize use of base64 2024-02-29 13:17:36 +01:00
6d33e721c4
[fix-auth-ct-eq] use consant time comparison for awsv4 signature verification 2024-02-29 13:07:18 +01:00
eaac4924ef
[fix-auth-ct-eq] use argon2 hashing and verification for admin/metrics token checking 2024-02-29 13:07:15 +01:00
02005055ae Merge pull request 'Mention deduplication and compression in features page' (#736) from doc-dedup into main
Reviewed-on: Deuxfleurs/garage#736
2024-02-28 12:49:26 +00:00
a294dd9473
[doc-dedup] reorder features, move no-RAFT down 2024-02-28 13:48:45 +01:00
947973982d
[doc-dedup] fix #rpc_bind_outgoing anchor in config page 2024-02-28 13:45:30 +01:00
dc995059aa
[doc-dedup] mention deduplication and compression in features page 2024-02-28 13:43:30 +01:00
10031a3a91 Merge pull request 'Split presigned signature verification + fix conditions' (#735) from fix-presigned into main
Reviewed-on: Deuxfleurs/garage#735
2024-02-28 11:38:00 +00:00
90cab5b8f2
[fix-presigned] add comments and reorganize 2024-02-28 12:24:21 +01:00
e9f759d4cb
[fix-presigned] presigned requests: allow x-amz-* query parameters to stand in for equivalent headers 2024-02-28 12:24:21 +01:00
a5e4bfeae9
[fix-presigned] write comments 2024-02-28 12:24:21 +01:00
4c1d42cc5f
[fix-presigned] add back anonymous request code path + refactoring 2024-02-28 12:24:21 +01:00
2efa9c5a1a
[fix-presigned] PostObject: verify X-Amz-Algorithm 2024-02-28 12:24:20 +01:00
a8cb8e8a8b
[fix-presigned] split presigned/normal signature verification 2024-02-28 12:24:13 +01:00
d0d95fd53f
[next-0.10] woodpecker: run debug pipeline on manual trigger 2024-02-27 10:13:09 +01:00
4b978b7533
Merge branch 'main' into next-0.10 2024-02-26 18:55:24 +01:00
911a83ea7d Merge pull request 'rewrite read_and_put_block as a series of steps with channels' (#734) from refactor-put into main
Reviewed-on: Deuxfleurs/garage#734
2024-02-26 17:52:45 +00:00
b76c0c102e
[refactor-put] add ordering tag to blocks being sent to storage nodes 2024-02-26 18:35:11 +01:00
babccd2ad3
[refactor-put] send several blocks in parallel to storage nodes 2024-02-26 18:22:37 +01:00
3fe94cc14f
[refactor-put] rewrite read_and_put_block as a series of steps with channels 2024-02-26 17:55:37 +01:00
ee2b0c8dda
[talk-capitoul] Add talk on 2024-02-09 at capitoul.org 2024-02-26 13:42:47 +01:00
17b55205aa Merge pull request 'doc: reverse-proxy.md: Added section on caddy-fs-s3' (#733) from jpds/garage:caddy-fileserver-browse-s3 into main
Reviewed-on: Deuxfleurs/garage#733
2024-02-26 09:56:09 +00:00
Jonathan Davies
3813e6c71d doc: reverse-proxy.md: Added section on caddy-fs-s3. 2024-02-26 00:54:33 +00:00
3692af7052
Merge branch 'main' into next-0.10 2024-02-23 18:28:05 +01:00
e399b60e25 Merge pull request 'GetObject: split out handle_get_full (small refactoring)' (#732) from split_getobject into main
Reviewed-on: Deuxfleurs/garage#732
2024-02-23 17:26:53 +00:00
d640102b76
[split_getobject] GetObject: split out handle_get_full 2024-02-23 18:14:50 +01:00
916c67ccf4
Merge branch 'main' into next-0.10 2024-02-23 16:50:34 +01:00
61758ce0f9 Merge pull request 'some refactoring on data read/write path' (#729) from refactor-block into main
Reviewed-on: Deuxfleurs/garage#729
2024-02-23 15:49:43 +00:00
6ee691e65f
[refactor-block] simplify some more 2024-02-23 12:50:10 +01:00
e9c42bca34
[refactor-block] add DataBlockStream type 2024-02-23 12:22:29 +01:00
cd1069c1d4
[refactor-block] refactor DataBlock and DataBlockPath 2024-02-23 12:15:52 +01:00
07c7895948
[refactor-block] simplify rpc_get_block 2024-02-23 11:54:40 +01:00
9b41f4ff20
[refactor-block] move read_stream_to_end to garage_net 2024-02-23 11:46:57 +01:00
93552b9275
[refactor-block] Remove redundant BlockStream type 2024-02-23 11:33:38 +01:00
81cebdd124
[next-0.10] fix build 2024-02-22 15:53:47 +01:00
59f61c966a
Merge branch 'main' into next-0.10 2024-02-22 15:45:45 +01:00
74d0c47f21 Merge pull request 'Add node-global lock for bucket/key operations (fix #723)' (#728) from lock-createbucket into main
Reviewed-on: Deuxfleurs/garage#728
2024-02-22 12:05:19 +00:00
cff702a951
[lock-createbucket] Add node-global lock for bucket/key operations (fix #723) 2024-02-22 12:28:21 +01:00
7e212e20e0 Merge pull request 'Minor typos & grammar fixes in docs' (#727) from hartraft/garage:docs-typo-fix into main
Reviewed-on: Deuxfleurs/garage#727
2024-02-22 09:26:08 +00:00
hartraft
00a5f14a7b Align admin endpoint port 2024-02-20 21:19:00 +01:00
hartraft
1a07c8dd54 Minor typos and grammar 2024-02-20 21:03:39 +01:00
292f4ff9cb Typo
Fix small typo on the getting started guide
2024-02-20 18:35:56 +00:00
75e591727d
[next-0.10] cluster node status metrics: report nodes of all active layout versions 2024-02-20 17:08:31 +01:00
643d1aabd8
Merge branch 'main' into next-0.10 2024-02-20 17:02:44 +01:00
885405d944 Merge pull request 'system metrics improvements' (#726) from peer-metrics into main
Reviewed-on: Deuxfleurs/garage#726
2024-02-20 15:35:12 +00:00
bcd571ef57
[peer-metrics] add documentation for new cluster status metrics 2024-02-20 14:59:04 +01:00
b868493da9
[peer-metrics] add basic cluster node status metrics (fix #545) 2024-02-20 14:50:24 +01:00
182a23cc12
[peer-metrics] refactor SystemMetrics to hold a reference to System 2024-02-20 14:20:58 +01:00
3cdf69f079
[peer-metrics] Add metrics for cluster health, like GetClusterHealth admin API 2024-02-20 13:50:45 +01:00
00d479358d
[peer-metrics] refactor/simplify SystemMetrics 2024-02-20 13:50:45 +01:00
203bb10035 Merge pull request 'Filter nodes Garage tries to connect to' (#719) from reconnect-only-current into main
Reviewed-on: Deuxfleurs/garage#719
2024-02-20 10:37:11 +00:00
e91576677e
[reconnect-only-current] filter nodes to reconnect to
do not try reconnecting to nodes received from consul/kubernetes
discovery if they are not currently in the layout
2024-02-20 11:07:10 +01:00
0b9859befa Merge pull request 'garage_net: retry connecting when new IP is learned' (#724) from networking-fixes into main
Reviewed-on: Deuxfleurs/garage#724
2024-02-19 17:37:01 +00:00
95e3a39b4d
[networking-fixes] small refactoring in garage_net peering 2024-02-19 18:12:21 +01:00
66fe893023
[networking-fixes] garage_net: retry connecting when new IP is learned 2024-02-19 18:12:21 +01:00
6bb34899f2 Merge pull request 'fixes to RPC networking' (#721) from networking-fixes into main
Reviewed-on: Deuxfleurs/garage#721
2024-02-19 11:44:05 +00:00
eab54b3798
[networking-fixes] add doc for rpc_bind_outgoing 2024-02-19 11:45:44 +01:00
b96f84b894
[networking-fixes] add option to bind outgoing RPC sockets (fix #638)
Thanks to yuka for the original patch.
2024-02-19 11:45:44 +01:00
f0bbad2db9
[networking-fixes] use rpc_public_addr in netapp's HelloMessage 2024-02-19 11:45:44 +01:00
b8217361c0 Merge pull request 'doc: fixes to index of configuration options' (#722) from doc-fixes into main
Reviewed-on: Deuxfleurs/garage#722
2024-02-19 10:45:14 +00:00
e73cb79e1e
[doc-fixes] configuration reference: fix typo and set block size in M 2024-02-19 11:42:06 +01:00
e54effec45
[doc-fixes] fixes to index of configuration options 2024-02-19 11:36:22 +01:00
eb4a6ce106
Merge branch 'main' into next-0.10 2024-02-15 14:06:34 +01:00
7be3f15e45 Merge pull request 'import Netapp code into Garage codebase' (#717) from import-netapp into main
Reviewed-on: Deuxfleurs/garage#717
2024-02-15 12:51:52 +00:00
125c662860
[import-netapp] move and rename FullMeshPeeringSrategy to PeeringManager 2024-02-15 12:15:35 +01:00
5766befb24
[import-netapp] fix tests 2024-02-15 12:15:33 +01:00
5ea24254a9
[import-netapp] import Netapp code into Garage codebase 2024-02-15 12:15:07 +01:00
a2ab275da8 Merge pull request 'Fix cargo warnings in Cargo.toml files' (#718) from fix-cargo-toml into main
Reviewed-on: Deuxfleurs/garage#718
2024-02-15 11:14:01 +00:00
1b0f167d2f
[fix-cargo-toml] fix cargo warnings in Cargo.toml files 2024-02-15 10:54:58 +01:00
cf2af186fc
Merge branch 'main' into next-0.10 2024-02-13 11:36:28 +01:00
823078b4cd Merge pull request 'small fixes to config/secrets handling' (#715) from fix-secrets-695 into main
Reviewed-on: Deuxfleurs/garage#715
2024-02-13 10:04:49 +00:00
ea09b483fe Merge pull request 'doc: mention warn and error as available log levels' (#716) from emilylange/garage:doc/mention-warn-error-log-levels into main
Reviewed-on: Deuxfleurs/garage#716
2024-02-13 08:13:23 +00:00
c86ac264cb
doc: mention warn and error as available log levels
For some users, this might be their first time being interacting with
the `env_logger` crate.
As such, they might not be aware that less verbose log levels exist.
Some might not want to log every incoming request, for example.

This commit also adds syntax hints to the code-fence for bash for better
syntax highlighting of that section, and repeats itself multiple times,
that `info` is, in fact, the default.

No changes to the recommendation of log levels were made.
2024-02-12 18:00:51 +01:00
bf283c9924
[fix-secrets-695] config: replace String by PathBuf for *_file 2024-02-12 15:36:43 +01:00
25e5738568
[fix-secrets-695] take into account rpc secret from file for cli commands (fix #695) 2024-02-12 10:42:17 +01:00
198188017c Merge pull request 'Implement header overriding in GetObject (fix #650)' (#713) from header-override-650 into main
Reviewed-on: Deuxfleurs/garage#713
2024-02-09 15:40:18 +00:00
02e98e2d10
[header-override-650] implement header overriding in GetObject (fix #650) 2024-02-09 15:58:46 +01:00
fe175fa8e2 Merge pull request 'garage block info: find blocks by prefix (fix #682)' (#712) from block-info-short-682 into main
Reviewed-on: Deuxfleurs/garage#712
2024-02-09 14:07:29 +00:00
3865080c35 Merge pull request 'Allow multi-character delimiters in List* (fix #692)' (#711) from multi-char-delimiter-692 into main
Reviewed-on: Deuxfleurs/garage#711
2024-02-09 13:38:17 +00:00
8da67b3aa2
[block-info-short-682] garage block info: find blocks by prefix (fix #682) 2024-02-09 14:35:53 +01:00
10bc2ead60
[multi-char-delimiter-692] allow multi-character delimiters in List* (fix #692) 2024-02-09 14:15:29 +01:00
0c7ce001c9 Merge pull request 'Fix & simplify CI using Woodpecker' (#706) from nix-improvements into main
Reviewed-on: Deuxfleurs/garage#706
2024-02-09 12:11:23 +00:00
f7ae966ed3
[nix-improvements] special case for Docker's "386" architecture 2024-02-09 12:49:17 +01:00
561fad0b44
[nix-improvements] get rid of Drone 2024-02-09 12:19:16 +01:00
1be75fbf4e
[nix-improvements] fix kaniko and manifest-tool 2024-02-09 11:46:46 +01:00
555ed75548
[nix-improvements] ci: check static as separate step 2024-02-09 11:36:51 +01:00
1c85e5e428
[nix-improvements] adapt woodpecker pipelines 2024-02-09 11:19:32 +01:00
d35d4599de
[nix-improvements] use kaniko and manifest-tools from nixpkgs, simplify 2024-02-09 11:15:52 +01:00
9900368380
[nix-improvements] modernize Nix infrastructure 2024-02-09 11:10:13 +01:00
e4a43bfd59 Merge pull request 'Upgrade toml, kube, k8s-openapi + code fixes' (#709) from dep-upgrade-202402 into main
Reviewed-on: Deuxfleurs/garage#709
2024-02-09 09:32:54 +00:00
5c63193d1d
[dep-upgrade-202402] fix shutdown issue introduced when upgrading hyper 2024-02-08 23:43:59 +01:00
bcbd15da84
[dep-upgrade-202402] cargo clippy fixes 2024-02-08 23:29:57 +01:00
ad5ce968d2
[dep-upgrade-202402] remove useless mut 2024-02-08 23:29:57 +01:00
c2e1e172d4
[dep-upgrade-202402] update toml, kube and k8s-openapi 2024-02-08 23:29:56 +01:00
8061bf5e1c Merge pull request 'Use only oxalica/rust-overlay toolchain and not nixpkgs' (#710) from oxalica-toolchain-only into main
Reviewed-on: Deuxfleurs/garage#710
2024-02-08 22:29:25 +00:00
8724aabdf5
[oxalica-toolchain-only] remove obsolete comment on toolchains 2024-02-08 23:23:27 +01:00
57024a2129
[oxalica-toolchain-only] remove custom toolchains from toolchains.nix 2024-02-08 23:21:00 +01:00
9e0b1dcf1c
[oxalica-toolchain-only] remove use of nixos rust toolchain 2024-02-08 19:10:26 +01:00
304a89c57b Merge pull request 'convert drone pipelines to woodpecker' (#708) from woodpecker into main
Reviewed-on: Deuxfleurs/garage#708
2024-02-08 17:46:00 +00:00
25c2f37667
[woodpecker] remove upgrade test on i386 2024-02-08 18:42:26 +01:00
4e62e86644
[woodpecker] disable docker image generation as auth is broken for now 2024-02-08 18:27:33 +01:00
8b6a44a53d
[woodpecker] convert drone pipelines to woodpecker 2024-02-08 18:24:52 +01:00
710680da15 Merge pull request 'update toolchain' (#705) from dep-upgrade-202402 into main
Reviewed-on: Deuxfleurs/garage#705
2024-02-08 14:58:18 +00:00
33e6db8b72
[dep-upgrade-202402] update rustc to 1.73 2024-02-08 12:33:09 +01:00
3a49f86073 Merge pull request 'Enable LTO for release builds using Nix' (#707) from lto-nix into main
Reviewed-on: Deuxfleurs/garage#707
2024-02-08 10:36:49 +00:00
2b92e8d7c6
[lto-nix] enable LTO for release builds using Nix 2024-02-08 10:22:23 +01:00
59930977e0 Merge pull request 'Cargo.toml: Enable full LTO in release builds and thin in dev builds.' (#704) from jpds/garage:release-build-lto into main
Reviewed-on: Deuxfleurs/garage#704
2024-02-07 16:37:02 +00:00
Jonathan Davies
620664ee9c Cargo.toml: Enable full LTO in release builds and thin in dev builds. 2024-02-07 16:11:27 +00:00
5d941e0100 Merge pull request 'Dependency upgrades: http, hyper, aws-sdk, smaller deps' (#703) from dep-upgrade-202402 into main
Reviewed-on: Deuxfleurs/garage#703
2024-02-07 14:59:40 +00:00
e011941964
[dep-upgrade-202402] refactor use of BodyStream 2024-02-07 15:32:51 +01:00
53746b59e5
[dep-upgrade-202402] slightly more explicit error management 2024-02-07 14:53:13 +01:00
a31d1bd496
[dep-upgrade-202402] fix obsolete DateTime::from_utc calls 2024-02-07 14:48:27 +01:00
e524e7a30d
[dep-upgrade-202402] rename BytesBody into ErrorBody for clarity 2024-02-07 14:45:52 +01:00
fe48d60d2b
[dep-upgrade-202402] refactor http listener code 2024-02-07 14:34:40 +01:00
22332e6c35
[dep-upgrade-202402] simplify/refactor GetObject 2024-02-05 20:26:33 +01:00
81ccd4586e
[dep-upgrade-202402] upgrade to http/hyper 1.x for tests 2024-02-05 19:57:35 +01:00
a22bd31920
[dep-upgrade-202402] migration to http/hyper 1.0 for k2v api 2024-02-05 19:27:12 +01:00
0bb5b77530
[dep-upgrade-202402] wip: port to http/hyper crates v1 2024-02-05 18:49:54 +01:00
6e69a1fffc
[dep-upgrade-202402] prepare migration to http/hyper 1.0 2024-02-05 14:44:12 +01:00
6e4229e29c
[dep-upgrade-202402] update aws-sdk dependencies 2024-02-05 14:02:45 +01:00
c0a7552015
[dep-upgrade-202402] upgrade easy dependencies 2024-02-05 13:58:23 +01:00
fe1af5d98b
[dep-upgrade-202402] refactor dependencies: move all as workspace deps 2024-02-05 13:02:02 +01:00
f65da26ae2
[dep-upgrade-202402] update dependency minor versions using cargo update 2024-02-05 12:26:31 +01:00
feeb076b7f Merge pull request 'Add FOSDEM'24 talk' (#702) from talk-fosdem-24 into main
Reviewed-on: Deuxfleurs/garage#702
2024-02-05 11:17:15 +00:00
fe37202f8f
[talk-fosdem-24] remove abstract.md 2024-02-05 12:16:09 +01:00
76e09c0472
[talk-fosdem-24] small change in talk 2024-02-01 11:43:21 +01:00
1d30cf36c8
[talk-fosdem-24] improve fosdem 24 talk 2024-01-30 14:27:39 +01:00
d45189e7b8 Merge pull request 'doc: fix some typos' (#696) from Armael/garage:typos into main
Reviewed-on: Deuxfleurs/garage#696
2024-01-28 14:15:47 +00:00
Armaël Guéneau
91a51dd3e8 doc: fix some typos 2024-01-27 14:51:15 +01:00
08a871390e Merge pull request 'convert_db: allow LMDB map size override' (#691) from zdenek.crha/garage:convert_db_lmdb_map_size into main
Reviewed-on: Deuxfleurs/garage#691
2024-01-24 08:19:45 +00:00
c7dad980b7
[talk-fosdem-24] remove geodistrib paper shot 2024-01-23 17:25:45 +01:00
c2541f280c
[talk-fosdem-24] WIP, write talk, modify lots of assets 2024-01-23 16:50:30 +01:00
0eef8a69f0 make all garage_db::Engine variants un-conditional
Having all Engine enum variants conditional causes compilation errors
when *none* of the DB engine features is enabled. This is not an issue
for full garage build, but affects crates that use garage_db as
dependency.

Change all variants to be present at all times. It solves compilation
errors and also allows us to better differentiate between invalid DB
engine name and engine with support not compiled in current binary.
2024-01-22 21:12:02 +01:00
4de7ac6023
FOSDEM'24 talk WIP 2024-01-22 18:52:14 +01:00
74e72fc996 convert_db: cleanup naming and comments for open overrides 2024-01-22 17:52:39 +01:00
7a3b863150 Merge pull request 'doc: add presentation at seed webinar 2024-01-12' (#693) from prez-seed-webinar-202401 into main
Reviewed-on: Deuxfleurs/garage#693
2024-01-22 13:49:08 +00:00
d2c40b12e8
doc/talks: refactor assets 2024-01-22 14:43:46 +01:00
cf0abbfe42
rm abstract 2024-01-22 14:33:48 +01:00
4b54e053df convert_db: prevent conversion between same input/output engine
Use optional DB open overrides for both input and output database.

Duplicating the same override flag for input/output would result in too
many, too long flags. It would be too costly for very rare edge-case
where converting between same DB engine, just with different flags.

Because overrides flags for different engines are disjoint and we are
preventing conversion between same input/ouput DB engine, we can have
only one set.

The override flag will be passed either to input or output, based on
engine type it belongs to. It will never be passed to both of them and
cause unwelcome surprise to user.
2024-01-18 17:57:56 +01:00
8527dd87cc convert_db: allow LMDB map size override 2024-01-17 21:20:34 +01:00
0263828560 Merge pull request 'Garage v0.9.1' (#689) from rel-v0.9.1 into main
Reviewed-on: Deuxfleurs/garage#689
2024-01-17 12:00:23 +00:00
ee57dd922b
Bump version to 0.9.1 2024-01-16 16:28:17 +01:00
9cfeea389a Merge pull request 'CLI help, comments & messages: make clear that full-length node ID = public key' (#688) from rename-public-key into main
Reviewed-on: Deuxfleurs/garage#688
2024-01-16 13:33:43 +00:00
82a29bf6e5
help, comments: make clear that full-length node ID = public key
Generally, avoid using the "public key" terminology
2024-01-16 14:04:11 +01:00
707d85f602 Merge pull request 'sync garage v0.9 with garage v0.8' (#657) from sync-08-09 into main
Reviewed-on: Deuxfleurs/garage#657
2024-01-16 11:33:27 +00:00
4c5be79b80 Garage v0.8.5
This minor release includes the following improvements and fixes:
 
 New features:
 
 - Configuration: make LMDB's `map_size` configurable and make `block_size` and `sled_cache_capacity` expressable as strings (such as `10M`) (#628, #630)
 
 - Add support for binding to Unix sockets for the S3, K2V, Admin and Web API servers (#640)
 
 - Move the `convert_db` command into the main Garage binary (#645)
 
 - Add support for specifying RPC secret and admin tokens as environment variables (#643)
 
 - Add `allow_world_readable_secrets` option to config file (#663, #685)
 
 Bug fixes:
 
 - Use `statvfs` instead of mount list to determine free space in metadata/data directories (#611, #631)
 
 - Add missing casts to fix 32-bit build (#632)
 
 - Fix error when none of the HTTP servers (S3/K2V/Admin/Web) is started and fix shutdown hang (#613, #633)
 
 - Add missing CORS headers to PostObject response (#609, #656)
 
 - Monitoring: finer histogram boundaries in Prometheus exported metrics (#531, #686)
 
 Other:
 
 - Documentation improvements (#641)
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCAAdFiEEwhSWp0+ubv79TiqUDkltFQljdr4FAmWmWvsACgkQDkltFQlj
 dr59rRAAiMGQpDUK0QqiCgrp1rcUhvtj3DsQEpT7F14Jo3I7bFDmONZolPbO8YAs
 VE4S4CBQogNH0lMQ6EvJYiBCxDWkxdVibKqDWOYJmUw3bZ6Ypn1eZIF0+Uf1TDI+
 C6CxYbyDQtqvm330K2Du2uOoGiIgm83b6jktK/0FtbAE2GWhtYmQwoelprAGH20i
 baaSfkZbBl8toUscakyhPVVSQ86BcVQ2jqL6Ofu4eQknjMRqCeAIQhMB2ikpiwBz
 hbTZ3x0EfJJqiHocfkTE3B3cPnDKuHDzxPRhLMB/olEpzoxaLJ2+tc0ziQdl06/F
 1c8nHM57L1IaDGKAkpcANnj3yVf3jfPqq9SEUNi+xSIWbvln91RvXU4RIB8hiZqa
 rqAHjDuys++3DoAUr/L4X233MWufVAEYT4B+jaPAv6ys35xhQwPAMJrA0OZEr+hE
 HQMPIG9uMDVjZ2QCgFYgC02kEqvxbsRSVnb0wjI7eoNOk0LKo154eJh1cOGd4Ibs
 yBTiIi1+Y7RCXNxcIHKlj5vMUHPBr2D8DVFj21kfZKUtMQ/8yScoiRC14ZR4J2xF
 IYe3aDm80l3tYgnPRVj4fOGiIPsqnZd4iazYKwj2cifB8tzYfyh5/9fv2aio8K5y
 0GAw4AoTtgg1hLMadbc3om7wy64IRaZzXjv59eYPEotZYdreVpM=
 =RVm8
 -----END PGP SIGNATURE-----
gpgsig -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCAAdFiEEwhSWp0+ubv79TiqUDkltFQljdr4FAmWmZKIACgkQDkltFQlj
 dr6wZRAA1XuOBax/7YsIix3ag0kjnwnGAx8wYaA+Jiojw2yv/+ePL6yGHcKA93lI
 SL8l5G06fTDgpbpfdVbgyRzGT2tmjrXvkygRWf2WMDZ9I+8WxUA2q8aWaEMiNmvd
 0cfzYi14TgX+O0wEbKeeqrXG0473/yThk5U1FNbdJd7rkJ4JzaOTthdk0LJLiEUG
 zQ/YIYx3FVFoVI0rdORb3HKzqYHjMAvpzNhEIeqkrpDEzplQ3jKvY+rYWQL3S9zE
 bHbkZPoT62OpJGMr04/1FUkB+ctsvUrM0CskruaSKWyD2M1xTo/Ug4jh5muVIcdJ
 hJis1/k5rV8JDTIkb6eAxKqfVzI+56yDxofT8rVF4JhvlzvXDLOa0uyDVyA8/6un
 ylWRzs2Mlj6/TbscmPjrdH8v2Lb0zjWxvXe2iYnHHfldWUlYuBtI6FZiG3uNjBCs
 7ns3xr4VOw13RM5auVkEQksIO6lru0kvH18GB3h6Msx67w2JUzl+PaNv8PdRtnmV
 0SfLUl1Nh8yT2h9qG6/3cDE9E1G/mjg8SgljoEe6ahs/BUZmLuTHTyBjf+P22ZbO
 DCITM3CwrV+y/aKnRdLvd6LOWFinUqMS8YvVSVqJh9vo9R+dt33LdBMdWjP4IYHF
 MbACe4FzeG3AXUcHB/mDCm7a2H2BFwzAovFy0SE639PfWBxNue0=
 =gzWq
 -----END PGP SIGNATURE-----

Merge tag 'v0.8.5' into sync-08-09

Garage v0.8.5

This minor release includes the following improvements and fixes:

New features:

- Configuration: make LMDB's `map_size` configurable and make `block_size` and `sled_cache_capacity` expressable as strings (such as `10M`) (#628, #630)

- Add support for binding to Unix sockets for the S3, K2V, Admin and Web API servers (#640)

- Move the `convert_db` command into the main Garage binary (#645)

- Add support for specifying RPC secret and admin tokens as environment variables (#643)

- Add `allow_world_readable_secrets` option to config file (#663, #685)

Bug fixes:

- Use `statvfs` instead of mount list to determine free space in metadata/data directories (#611, #631)

- Add missing casts to fix 32-bit build (#632)

- Fix error when none of the HTTP servers (S3/K2V/Admin/Web) is started and fix shutdown hang (#613, #633)

- Add missing CORS headers to PostObject response (#609, #656)

- Monitoring: finer histogram boundaries in Prometheus exported metrics (#531, #686)

Other:

- Documentation improvements (#641)
2024-01-16 12:12:27 +01:00
083e982f5f Merge pull request 'Garage v0.8.5' (#687) from rel-0.8.5 into main-0.8.x
Reviewed-on: Deuxfleurs/garage#687
2024-01-16 10:30:54 +00:00
50643e61bf
Bump version to 0.8.5 2024-01-16 10:47:33 +01:00
a6421ee5a5 Merge pull request 'monitoring: finer histogram boundaries in prometheus metrics (fix #531)' (#686) from fix-531 into main-0.8.x
Reviewed-on: Deuxfleurs/garage#686
2024-01-15 16:44:58 +00:00
993ce74976 Merge pull request '0.8.x: config: refactor secret sourcing' (#685) from secret-sourcing into main-0.8.x
Reviewed-on: Deuxfleurs/garage#685
2024-01-15 16:41:50 +00:00
f512609123
monitoring: finer histogram boundaries in prometheus metrics (fix #531) 2024-01-15 17:33:35 +01:00
97bae7213a
config: additional tests for secret sourcing 2024-01-15 17:30:30 +01:00
7228695ee2
config: refactor secret sourcing 2024-01-15 17:18:46 +01:00
ee7fe27d3d Merge pull request 'Add allow_world_readable_secrets option to config file' (#663) from PicNoir/garage:nin/world-readable-conf-file into main-0.8.x
Reviewed-on: Deuxfleurs/garage#663
2024-01-15 15:20:16 +00:00
d91a1de731 Merge pull request 'fix typo in peertube doc' (#617) from Lapineige/garage:main into main
Reviewed-on: Deuxfleurs/garage#617
2024-01-11 11:19:42 +00:00
db48dd3d6c
bump crate versions to 0.10.0 2024-01-11 12:05:51 +01:00
8a6ec1d611 Merge pull request 'NLnet task 3' (#667) from nlnet-task3 into next-0.10
Reviewed-on: Deuxfleurs/garage#667
2024-01-11 10:58:08 +00:00
723e56b37f Merge pull request 'Jepsen testing (NLnet task 3 subtask 1)' (#544) from jepsen into main
Reviewed-on: Deuxfleurs/garage#544
2024-01-11 10:52:12 +00:00
60f0bd03b6
doc: add talk for SEED webinar 2024-01-11 11:40:44 +01:00
fa9247f11b jepsen: updated results, confirming that task3 works 2023-12-14 16:23:48 +01:00
0041b013a4
layout: refactoring and fix in layout helper 2023-12-11 16:09:22 +01:00
adccce1145
layout: refactor/fix bad while loop 2023-12-11 15:45:14 +01:00
85b5a6bcd1
fix some clippy lints 2023-12-11 15:31:47 +01:00
e4f493b481
table: remove redundant tracing in insert_many 2023-12-11 14:57:42 +01:00
f8df90b79b
table: fix insert_many to not send duplicates 2023-12-08 14:54:11 +01:00
4dbf254512
layout: refactoring, merge two files 2023-12-08 14:15:52 +01:00
64a6e557a4
rpc helper: small refactorings 2023-12-08 12:18:12 +01:00
5dd200c015
layout: move block_read_nodes_of to rpc_helper to avoid double-locking
(in theory, this could have caused a deadlock)
2023-12-08 12:02:24 +01:00
063294dd56
layout version: refactor get_node_zone 2023-12-08 11:50:58 +01:00
7f2541101f
cli: improvements to the layout commands when multiple layouts are live 2023-12-08 11:24:23 +01:00
91b874c4ef
rpc: fix system::health 2023-12-08 10:36:37 +01:00
431b28e0cf
fix build with discovery features 2023-12-07 15:15:59 +01:00
9cecea64d4
layout: allow sync update tracker to progress with only quorums 2023-12-07 14:51:20 +01:00
aa59059a91
layout cli: safer skip-dead-nodes command 2023-12-07 11:56:14 +01:00
d90de365b3
table sync: use write quorums to report global success or failure of sync 2023-12-07 11:16:10 +01:00
95eb13eb08
rpc: refactor result tracking for quorum sets 2023-12-07 10:57:21 +01:00
c8356a91d9
layout updates: fix the set of nodes among which minima are calculated 2023-12-07 10:30:26 +01:00
a8b0e01f88 Merge pull request 'OpenAPI specification of admin APIv1' (#672) from api-v1 into main
Reviewed-on: Deuxfleurs/garage#672
2023-11-29 15:42:46 +00:00
8088690650
fix the doc 2023-11-28 16:18:28 +01:00
c04dd8788a
admin: more info in admin GetClusterStatus 2023-11-28 14:25:04 +01:00
ffa659433d Merge pull request 'Doc: fix db_engines section and improve config reference' (#674) from fix-doc-db-engine into main
Reviewed-on: Deuxfleurs/garage#674
2023-11-28 12:03:46 +00:00
cfa5550cb2 doc: move replication_mode to top of configuration page reference 2023-11-28 11:58:27 +01:00
939d1f2e17 doc: improve navigation in configuration reference 2023-11-28 11:53:26 +01:00
1f6efe57be doc: update the db_engine section 2023-11-28 11:33:31 +01:00
539af6eac4
rpc helper: write comments + small refactoring of tracing 2023-11-28 11:12:39 +01:00
3908619eac
add ClusterHealthReport endpoint to the API 2023-11-28 09:34:01 +01:00
c539077d30
cli: remove historic layout info from status 2023-11-27 16:22:27 +01:00
11e6fef93c
cli: add layout history and layout assume-sync commands 2023-11-27 16:22:25 +01:00
539a920313
cli: show when nodes are draining metadata 2023-11-27 13:18:59 +01:00
78362140f5
rpc: update system::health to take into account write sets for all partitions 2023-11-27 12:10:21 +01:00
d6d239fc79
block manager: read_block using old layout versions if necessary 2023-11-27 11:52:57 +01:00
68d23cccdf
disable int64 finally for now 2023-11-23 10:20:36 +01:00
9f1043586c
set layout version as required 2023-11-23 10:16:16 +01:00
1caa6e29e5
capacity is int64 2023-11-23 10:02:41 +01:00
814b3e11d4
fix query parameters for keys 2023-11-23 08:50:10 +01:00
2d37e7fa39
convert showsecretkey from bool to enum 2023-11-22 21:05:36 +01:00
4f473f43c9
Change how query parameters are handled 2023-11-22 20:39:38 +01:00
3684c29ad0
handle key changes 2023-11-22 18:14:38 +01:00
0d415f42ac
Port GetKeyInfo by adding showSecretKey query param 2023-11-22 18:05:11 +01:00
20b3afbde4
Port layout endpoints 2023-11-22 17:49:51 +01:00
e3cd6ed530
port GetLayout and AddLayout 2023-11-22 15:24:30 +01:00
9b24d7c402
Upgrade GetNodes 2023-11-22 14:25:04 +01:00
36bd21a148 Merge pull request 'Allow 0 as a part number marker' (#670) from asonix/garage:main into main
Reviewed-on: Deuxfleurs/garage#670
2023-11-22 10:33:31 +00:00
d1d1940252
Health info message now advertises API v1 2023-11-22 09:28:50 +01:00
c63b446989
skeleton for api v1 2023-11-22 08:58:09 +01:00
92fd899fb6 Allow 0 as a part number marker 2023-11-21 17:39:51 -06:00
92dd2bbe15 jepsen: nlnet task3a seems to fix things 2023-11-16 18:09:13 +01:00
3ecd14b9f6
table: implement write sets for insert_many 2023-11-16 16:41:45 +01:00
22f38808e7
rpc_helper: don't use tokio::spawn for individual requests 2023-11-16 16:34:01 +01:00
707442f5de
layout: refactor digests and add "!=" assertions before epidemic bcast 2023-11-16 13:51:40 +01:00
ad5c6f779f
layout: split helper in separate file; more precise difference tracking 2023-11-16 13:26:43 +01:00
18e5811159
jepsen: add patch and use more complete names 2023-11-16 12:57:21 +01:00
d4df03424f
layout: fix test 2023-11-15 15:56:57 +01:00
33c8a489b0
layou: implement ack locking 2023-11-15 15:40:44 +01:00
393c4d4515
layout: add helper for cached/external values to centralize recomputation 2023-11-15 14:20:50 +01:00
65066c7064
layout: wip cache global mins 2023-11-15 13:28:30 +01:00
acd49de9f9
rpc: fix write set quorums 2023-11-15 13:07:42 +01:00
46007bf01d
integration test: print stdout and stderr on subcommand crash 2023-11-15 12:56:52 +01:00
b3e729f4b8
layout history merge: rm invalid versions when valid versions are added 2023-11-15 12:15:58 +01:00
7ef2c23120
layout: fix test 2023-11-14 15:45:01 +01:00
90e1619b1e
table: take into account multiple write sets in inserts 2023-11-14 15:40:46 +01:00
3b361d2959
layout: prepare for write sets 2023-11-14 14:28:16 +01:00
866196750f
system: add todo wrt new layout 2023-11-14 13:36:58 +01:00
83a11374ca
layout: fixes in schema 2023-11-14 13:29:26 +01:00
1aab1f4e68
layout: refactoring of all_nodes 2023-11-14 13:12:32 +01:00
8e292e06b3
layout: some refactoring of nongateway nodes 2023-11-14 12:48:38 +01:00
9a491fa137
layout: fix test 2023-11-11 13:10:59 +01:00
df24bb806d
layout/sync: fix bugs and add tracing 2023-11-11 12:44:27 +01:00
ce89d1ddab
table sync: adapt to new layout history 2023-11-11 12:08:32 +01:00
df36cf3099
layout: add helpers to LayoutHistory and prepare integration with Table 2023-11-09 16:32:31 +01:00
9d95f6f704
layout: fix tracker bugs 2023-11-09 15:52:45 +01:00
bad7cc812e
layout admin: add missing calls to update_hash 2023-11-09 15:42:10 +01:00
03ebf18830
layout: begin managing the update tracker values 2023-11-09 15:31:59 +01:00
94caf9c0c1
layout: separate code path for synchronizing update trackers only 2023-11-09 14:53:34 +01:00
bfb1845fdc
layout: refactor to use a RwLock on LayoutHistory 2023-11-09 14:12:05 +01:00
19ef1ec8e7
layout: more refactoring 2023-11-09 13:34:14 +01:00
8a2b1dd422
wip: split out layout management from System into separate LayoutManager 2023-11-09 12:55:36 +01:00
523d2ecb95
layout: use separate CRDT for staged layout changes 2023-11-09 11:19:43 +01:00
1da0a5676e
bump garage protocol version tag to 0x000A (0.10) 2023-11-08 19:30:58 +01:00
8dccee3ccf
cluster layout: adapt all uses of ClusterLayout to LayoutHistory 2023-11-08 19:28:36 +01:00
fe9af1dcaa
WIP: garage_rpc: store layout version history 2023-11-08 17:49:06 +01:00
4a9c94514f
avoid using layout_watch in System directly 2023-11-08 16:41:00 +01:00
12d1dbfc6b
remove Ring and use ClusterLayout everywhere 2023-11-08 15:41:24 +01:00
0962313ebd
garage_rpc: reorder functions in layout.rs 2023-11-08 13:13:04 +01:00
f83fa02193 Add allow_world_readable_secrets option to config file
Sometimes, the secret files permissions checks gets in the way. It's
by no mean complete, it doesn't take the Posix ACLs into account among
other things. Correctly checking the ACLs would be too involving (see
Deuxfleurs/garage#658 (comment))
and would likely still fail in some weird chmod settings.

We're adding a new configuration file key allowing the user to disable
this permission check altogether.

The (already existing) env variable counterpart always take precedence
to this config file option. That's useful in cases where the
configuration file is static and cannot be easily altered.

Fixes Deuxfleurs/garage#658

Co-authored-by: Florian Klink <flokli@flokli.de>
2023-10-26 18:25:13 +02:00
f4d3905d15 Merge pull request 'nix: add clang to flake.nix and shell.nix' (#664) from add-clang into main
Reviewed-on: Deuxfleurs/garage#664
2023-10-26 09:25:53 +00:00
a0fa50dfcd Merge pull request 's3 api: refactoring and bug fix in ListObjects' (#655) from fix-list-objects into main
Reviewed-on: Deuxfleurs/garage#655
2023-10-26 09:22:47 +00:00
d50fa2a562
nix: add clang to flake.nix and shell.nix 2023-10-26 11:19:22 +02:00
4b3dee2ca3 Merge pull request 's3 api: add missing CORS headers to PostObject responses (fix #609)' (#656) from fix-cors-post-object into main-0.8.x
Reviewed-on: Deuxfleurs/garage#656
2023-10-26 09:17:14 +00:00
5b1f50be65 jepsen: testing 2023-10-25 14:43:24 +02:00
9df7fa0bcd jepsen: use 7 nodes 2023-10-25 14:04:39 +02:00
fd85010a40 jepsen: failures with set2 test in --scenario r 2023-10-25 12:13:27 +02:00
cfbfa09d24 jepsen: fix set2 test omg finally this is so stupid 2023-10-25 11:50:16 +02:00
db921cc05f jepsen: reconfigure nemesis + add db nemesis 2023-10-25 11:41:34 +02:00
4fa2646a75 jepsen: got a failure with set1 2023-10-24 17:45:22 +02:00
d7ab2c639e jepsen: fix nemesis to actually generate many operations 2023-10-24 16:39:50 +02:00
d13bde5e26 jepsen: set1 and set2 don't fail anymore ?? 2023-10-24 15:44:05 +02:00
75d5d08ee1 Merge pull request 'Ensure increasing version timestamps when writing new object versions' (#543) from increasing-timestamps into main
Reviewed-on: Deuxfleurs/garage#543
2023-10-24 10:07:16 +00:00
d2c365767b jepsen: more testing 2023-10-24 11:39:45 +02:00
fb6c9a1243 jepsen: update readme 2023-10-20 15:55:09 +02:00
9030c1eef8 jepsen: code path for nemesis final generator 2023-10-20 15:53:46 +02:00
654775308e jepsen: add cluster reconfiguration nemesis 2023-10-20 15:48:37 +02:00
f5b0972781 jepsen: register crdt read-after-write is fixed with deleteobject patch 2023-10-20 15:00:10 +02:00
c82d91c6bc DeleteObject: always insert a deletion marker with a bigger timestamp than everything before 2023-10-20 13:56:35 +02:00
8686cfd0b1 s3 api: also ensure increasing timestamps for create_multipart_upload 2023-10-20 13:37:37 +02:00
d148b83d4f jepsen: reg2 failure seems to happen only with deleteobject 2023-10-20 13:36:48 +02:00
c6cde1f143 remove now-unused key parameter in check_quotas 2023-10-20 13:20:47 +02:00
4b93ce179a jepsen: errors in reg2 workload under investigation 2023-10-20 12:56:55 +02:00
4ba18ce9cc jepsen: wip checker for register-like behavior 2023-10-20 12:13:11 +02:00
ac04934dae s3 api: add missing CORS headers to PostObject responses (fix #609) 2023-10-20 10:37:48 +02:00
ef662822c9 jepsen: fix the list-objects call (?) 2023-10-19 23:40:55 +02:00
da8b170748 jepsen: investigating listobjects error 2023-10-19 16:45:24 +02:00
58b0ee1b1a list objects: prettyness and add asserts 2023-10-19 15:26:17 +02:00
158dc17a06 listobjects: fix panic if continuation token is an empty string 2023-10-19 15:08:47 +02:00
74e50edddd jepsen: refactoring 2023-10-19 14:34:19 +02:00
0215b11402 Merge pull request 'Add support for specifying rpc_secret_file, metrics_token_file and admin_token_file using environment variables' (#643) from networkException/garage:token-file-env into main-0.8.x
Reviewed-on: Deuxfleurs/garage#643
2023-10-19 09:33:12 +00:00
8599051c49
garage: support specifying token / secret as environment variables
this patch adds support for specifying the `rpc_secret_file`,
`metrics_token_file` and `admin_token_file` as environment variables.
2023-10-19 03:39:02 +02:00
4a19ee94bb
garage: fix admin-token description 2023-10-19 03:31:50 +02:00
c99cb58d71
util: move reading secret file into seperate helper
this patch moves the logic to read a secret file (and check for correct
permissions) from `secret_from_file` into a new `read_secret_file`
helper.
2023-10-19 03:29:48 +02:00
5feb6a1f64
docs: add documentation for specifying token / secret file as environment variables 2023-10-19 03:28:44 +02:00
b3bf16ee27 make jepsen test more robust: handle errors and timeouts, fixed access key 2023-10-18 17:51:34 +02:00
d146cdd5b6 cargo fmt 2023-10-18 16:38:26 +02:00
3d6ed63824 check_quotas: avoid re-fetching object from object table 2023-10-18 16:36:48 +02:00
45b0453d0f Ensure increasing version timestamps in PutObject 2023-10-18 16:31:50 +02:00
ddd3de7fce refactor jepsen code 2023-10-18 16:30:45 +02:00
84d43501ce refactor jepsen setup logic 2023-10-18 15:34:12 +02:00
012ade5d4b jepsen: update jepsen and fix garage key info 2023-10-18 14:06:32 +02:00
ef5ca86dfc jepsen: update to garage 0.9.0 2023-10-18 14:01:18 +02:00
9ec4cca334 reformatting 2023-10-18 12:03:12 +02:00
18ee8efb5f Check read-after-write property for sets 2023-10-18 12:03:12 +02:00
55eb4e87c4 set tests with independant tests together 2023-10-18 12:03:11 +02:00
0bb1577ae1 two set workloads with different checkers 2023-10-18 12:03:11 +02:00
6eb26be548 Add garage set test (this one works :p) 2023-10-18 12:03:11 +02:00
eb86eaa6d2 refactor jepsen test 2023-10-18 12:03:11 +02:00
80d7b7d858 remove useless files 2023-10-18 12:03:11 +02:00
93a7132b4c the fix for increasing timestamps does not make things linearizable 2023-10-18 12:03:11 +02:00
dc5245ce65 even without nemesis, s3 get/put/delete is not linearizable (is this normal?) 2023-10-18 12:03:11 +02:00
70c1d3db46 better match exceptions 2023-10-18 12:03:11 +02:00
bc11701999 jepsen: s3 gets and puts 2023-10-18 12:03:11 +02:00
ca4cc7e44f jepsen connects to vagrant vms 2023-10-18 12:03:11 +02:00
17ebb65273 jepsen ssh into containers seem to work ? 2023-10-18 12:03:11 +02:00
7011b71fbd jepsen: wip 2023-10-18 12:03:11 +02:00
a5e8ffeb63 Merge pull request 'use mold linker when invoking cargo manually (not in nix build scripts)' (#646) from mold-linker into main
Reviewed-on: Deuxfleurs/garage#646
2023-10-18 10:02:34 +00:00
b53510c5b7 Merge pull request 'fix compilation on macos' (#654) from trinity-1686a/garage:fix-macos-compilation into main
Reviewed-on: Deuxfleurs/garage#654
2023-10-16 09:33:33 +00:00
c7f5dcd953 fix compilation on macos
fsblkcnt_t is ony 32b there, so we have to do an additional cast
2023-10-15 17:57:27 +02:00
d8263fdf92 Merge pull request 'documentation updates for v0.9.0' (#647) from doc-updates into main
Reviewed-on: Deuxfleurs/garage#647
2023-10-11 12:57:37 +00:00
d24aaba697 doc: update quick start and real world for v0.9.0 2023-10-11 14:49:54 +02:00
b571dcd811 doc: updates to the "migrating to v0.9" page 2023-10-10 15:43:26 +02:00
e6df7089a1 Merge pull request 'Garage v0.9' (#473) from next into main
Reviewed-on: Deuxfleurs/garage#473
2023-10-10 13:28:28 +00:00
952c9570c4 bump version to v0.9.0 2023-10-10 14:08:11 +02:00
3d7892477d convert_db: fix build 2023-10-10 14:06:25 +02:00
d4932c31ea Merge branch 'main' into next 2023-10-10 13:57:21 +02:00
d3fffd30dc use mold linker when invoking cargo manually (not in nix build scripts) 2023-10-10 13:56:48 +02:00
e75fe2157d Merge pull request 'Move convert_db command into main garage binary' (#645) from convert-db-main-binary into main
Reviewed-on: Deuxfleurs/garage#645
2023-10-10 11:42:14 +00:00
2d5d7a7031 Move convert_db command into main garage binary 2023-10-10 12:13:15 +02:00
0c431b0c03 admin api: increased compatibility for v0/ endpoints 2023-10-05 16:56:13 +02:00
1c13135f25 admin api: remove broken GET /v0/key router rule 2023-10-05 16:27:29 +02:00
2448eb7713 upgrade doc: fixes and precisions 2023-10-05 15:29:55 +02:00
6790e24f5a Add migration to v0.9 guide 2023-10-05 15:20:48 +02:00
9ccc1d6f4a move upgrade test to release build 2023-10-05 10:42:10 +02:00
920dec393a cli: more precise doc comment 2023-10-04 10:44:42 +02:00
2e656b541b Merge branch 'main' into next 2023-10-03 18:40:37 +02:00
1243db87f2 Merge pull request 'Add support for binding to unix domain sockets' (#640) from networkException/garage:unix-sockets into main
Reviewed-on: Deuxfleurs/garage#640
2023-10-03 16:23:02 +00:00
6f8a87814b
doc: add documentation for specifying unix socket paths 2023-10-03 17:56:34 +02:00
7907a09acc
api: allow custom unix bind mode and use 0o220 for admin server 2023-10-03 17:31:40 +02:00
16aa418e47 Merge pull request 'doc: update endpoint_url documentation' (#641) from flokli/garage:aws-endpoint-url into main
Reviewed-on: Deuxfleurs/garage#641
2023-10-02 14:30:53 +00:00
cb359b4434 doc: update endpoint_url documentation
Since `awscli` `>=1.29.0` or `>=2.13.0` it is now possible to use the
`AWS_ENDPOINT_URL` environment variable, or the `endpoint_url` config
key to override the endpoint URL. This means, the aws bash function to
wrap with --endpoint-url is not necessary anymore. Update invocations to
reflect that.

https://docs.aws.amazon.com/sdkref/latest/guide/feature-ss-endpoints.html
https://github.com/aws/aws-cli/issues/4454#issuecomment-1626116607
2023-10-02 17:16:11 +03:00
8ec6a53b35
everywhere: support unix sockets when binding in various places
this patch implements binding to paths as a unix socket for generic
server and web server.
2023-09-29 18:57:44 +02:00
7353038a64
config: allow using paths for unix domain sockets in various places
this patch updates the config format to also allow paths in bind
addresses for unix domain sockets.

this has been added to all apis except rpc.
2023-09-29 18:38:30 +02:00
10195f1567
util: add helper sum type for unix and tcp socket addresses
this patch introduces a new sum type that can represent either a
tcp socket address or a unix domain socket path.
2023-09-29 18:37:36 +02:00
6086a3fa07
cargo: add hyperlocal as a dependency 2023-09-29 18:37:12 +02:00
9ac1d5be0e add upgrade test for garage 0.8 -> 0.9 2023-09-27 14:57:37 +02:00
897cbf2c27 actually update rmp-serde to 1.1.2 for both garage and netapp dependency (fix #629) 2023-09-27 13:13:00 +02:00
ad82035b98 Merge branch 'main' into next 2023-09-27 13:11:52 +02:00
aa7eadc799 Merge pull request 'New layout: fixes and UX improvements' (#634) from new-layout-ux into next
Reviewed-on: Deuxfleurs/garage#634
2023-09-27 09:04:32 +00:00
0e5925fff6 layout doc: reformulate 2023-09-22 16:14:47 +02:00
8d07888fa2 layout doc: write explanations for bizarre scenarios 2023-09-22 16:07:46 +02:00
405aa42b7d layout doc: update old text 2023-09-22 10:06:31 +02:00
b4a0e636d8 new layout doc: add examples of unexpected layout, to explain 2023-09-22 09:49:07 +02:00
1d986bd889 Merge pull request 'Refactor db transactions and add on_commit for table.queue_insert' (#637) from k2v-indices-lmdb into next
Reviewed-on: Deuxfleurs/garage#637
2023-09-21 14:03:35 +00:00
0635250b2b garage_table/queue_insert: delay worker notification to after transaction commit (fix #583) 2023-09-21 15:37:28 +02:00
f97168f805 garage_db: refactor transactions and add on_commit mechanism 2023-09-21 15:35:31 +02:00
3ecc17f8c5 new layout: use deterministic randomness for reproducible results 2023-09-21 11:21:35 +02:00
3a0e074047 Merge pull request 'prez-ocp' (#636) from prez-ocp into main
Reviewed-on: Deuxfleurs/garage#636
2023-09-21 08:15:10 +00:00
95ae09917b add ocp2023 presentation 2023-09-19 14:02:07 +02:00
a7ababb5db doc: update sticker 2023-09-18 16:40:06 +02:00
013b026d56 update cargo.nix 2023-09-18 12:18:56 +02:00
0088599f52 new layout: fix clippy lints 2023-09-18 12:17:07 +02:00
749b4865d0 new layout: improve display and fix comments 2023-09-18 12:07:45 +02:00
015ccb39aa new layout: make zone_redundancy optionnal (if not set, is maximum) 2023-09-18 11:59:08 +02:00
2e229d4430 new layout: improve output display 2023-09-12 17:24:51 +02:00
be1a16b42b Merge pull request 'Fix multiple shutdown issues' (#633) from fix-shutdown into main
Reviewed-on: Deuxfleurs/garage#633
2023-09-12 12:54:50 +00:00
91e764a2bf fix hang on shutdown 2023-09-12 14:35:48 +02:00
aa79810596 Fix error when none of S3/K2V/WEB/ADMIN server is started (fix #613) 2023-09-12 14:35:19 +02:00
fd7d8fec59 Merge branch 'main' into next 2023-09-11 23:09:20 +02:00
143a349f55 Merge pull request 'fix 32-bit build' (#632) from fix-32bit into main
Reviewed-on: Deuxfleurs/garage#632
2023-09-11 21:08:26 +00:00
9cfe55ab60 fix 32-bit build 2023-09-11 20:01:29 +02:00
51abbb02d8 Merge branch 'main' into next 2023-09-11 20:00:02 +02:00
2548a247f2 Merge pull request 'use statvfs instead of mount list to determine free data/meta space (fix #611)' (#631) from fix-free-space into main
Reviewed-on: Deuxfleurs/garage#631
2023-09-11 17:29:23 +00:00
d5bb50d738 use statvfs instead of mount list to determine free data/meta space (fix #611) 2023-09-11 19:08:24 +02:00
fc635f7072 Merge pull request 'make lmdb's map_size configurable (fix #628)' (#630) from configurable-map-size into main
Reviewed-on: Deuxfleurs/garage#630
2023-09-11 16:48:14 +00:00
f8b3883611 config: make block_size and sled_cache_capacity expressable as strings 2023-09-11 18:34:59 +02:00
51b9731a08 make lmdb's map_size configurable (fix #628) 2023-09-11 18:03:44 +02:00
ad6b1cc0be Merge branch 'main' into next 2023-09-11 13:14:18 +02:00
7228fbfd4f Merge pull request 'multi-hdd support (fix #218)' (#625) from multihdd into next
Reviewed-on: Deuxfleurs/garage#625
2023-09-11 10:52:01 +00:00
ba7ac52c19 block repair: simpler/more robust iterator progress calculation 2023-09-11 12:31:34 +02:00
9526328d38 scrub: clear saved checkpoint when canceling scrub 2023-09-11 12:10:48 +02:00
7f9ba49c71 block manager: remove data_dir field 2023-09-11 11:57:36 +02:00
de5d792181 block manager: fix indentation (why not detected by cargo fmt?) 2023-09-11 11:52:57 +02:00
be91ef6294 block manager: fix bug where rebalance didn't delete old copies 2023-09-07 16:04:03 +02:00
2657b5c1b9 block manager: fix bugs 2023-09-07 15:30:56 +02:00
eb972a8422 doc: update multi-hdd section 2023-09-07 14:48:36 +02:00
2f112ac682 correct free data space accounting for multiple data dirs on same fs 2023-09-07 14:42:20 +02:00
6a067e30ee doc: documentation of rebalance repair 2023-09-07 13:49:12 +02:00
6b008b5bd3 block manager: add rebalance operation to rebalance multi-hdd setups 2023-09-07 13:44:11 +02:00
6595efd82f Document multi-hdd support 2023-09-07 13:23:02 +02:00
bca347a1e8 doc: update page on upgradin clusters 2023-09-07 12:52:44 +02:00
99ed18350f block manager: refactor and fix monitoring/statistics 2023-09-07 12:41:36 +02:00
f38a31b330 block manager: avoid incorrect data_dir configs and avoid losing files 2023-09-06 17:49:30 +02:00
e30865984a block manager: scrub checkpointing 2023-09-06 16:35:28 +02:00
55c514999e block manager: fixes in layout 2023-09-06 16:35:28 +02:00
a44f486931 block manager: refactoring & increase max worker count to 8 2023-09-06 16:35:28 +02:00
3a74844df0 block manager: fix dir_not_empty 2023-09-06 16:35:28 +02:00
93114a9747 block manager: refactoring 2023-09-06 16:35:28 +02:00
fd00a47ddc table queue: increase batch size 2023-09-06 16:35:28 +02:00
1b8c265c14 block manager: get rid of check_block_status 2023-09-06 16:35:28 +02:00
3199cab4c8 update cargo.nix 2023-09-06 16:35:28 +02:00
a09f86729c block manager: move blocks in write_block if necessary 2023-09-06 16:35:28 +02:00
887b3233f4 block manager: use data paths from layout 2023-09-06 16:35:28 +02:00
6c420c0880 block manager: multi-directory layout computation 2023-09-06 16:35:28 +02:00
71c0188055 block manager: skeleton for multi-hdd support 2023-09-06 16:35:28 +02:00
4b4f2000f4 lifecycle: fix SkipBucket bug 2023-09-06 16:34:07 +02:00
5f86b48f97 Merge pull request 'Revert netapp to 0.5.2 to avoid rmp-serde upgrade that breaks things' (#627) from hold-netapp-0.5.2 into main
Reviewed-on: Deuxfleurs/garage#627
2023-09-05 22:08:40 +00:00
51eac97260 update version to 0.8.4 2023-09-05 23:28:12 +02:00
e78566591b Revert netapp update, hold to version 0.5.2 that uses rmp-serde 0.15 2023-09-05 23:23:23 +02:00
3f461d8891 Merge pull request 'object lifecycles (fix #309)' (#620) from bucket-lifecycle into next
Reviewed-on: Deuxfleurs/garage#620
2023-09-04 09:45:10 +00:00
8e0c020bb9 lifecycle worker: correct small clippy lints 2023-09-04 11:33:44 +02:00
1cdc321e28 lifecycle worker: don't get stuck on non-existent bucket 2023-08-31 11:36:30 +02:00
f579d6d9b4 lifecycle worker: fix potential inifinite loop 2023-08-31 11:29:54 +02:00
a00a52633f lifecycle worker: add log message when starting 2023-08-31 11:25:14 +02:00
adbf5925de lifecycle worker: use queue_insert and process objects in batches 2023-08-31 11:19:26 +02:00
1cfcc61de8 lifecycle worker: mitigate potential bugs + refactoring 2023-08-31 00:28:37 +02:00
be03a4610f s3api: remove redundant serde rename attribute 2023-08-31 00:00:26 +02:00
b2f679675e lifecycle worker: take into account disabled rules 2023-08-30 23:52:09 +02:00
5fad4c4658 update cargo.nix 2023-08-30 23:47:42 +02:00
01c327a07a lifecycle worker: avoid building chrono's serde feature 2023-08-30 23:46:15 +02:00
f0a395e2e5 s3 bucket apis: remove redundant call 2023-08-30 23:39:28 +02:00
d94f1c9178 reference manual: remove obsolete caveat about multipart uploads 2023-08-30 23:27:02 +02:00
5c923d48d7 reference manual: document support for lifecycle configuration 2023-08-30 23:24:28 +02:00
a1d57283c0 bucket_table: bucketparams::new doesn't need to be pub 2023-08-30 20:07:14 +02:00
d2e94e36d6 lifecycle config: add missing line in merge() and remove tracing 2023-08-30 20:05:53 +02:00
75ccc5a95c lifecycle config: store date as given, try to debug 2023-08-30 20:02:07 +02:00
7200954318 lifecycle worker: add logging 2023-08-30 14:54:52 +02:00
0f1849e1ac lifecycle worker: launch with the rest of Garage 2023-08-30 14:51:08 +02:00
da8b224e24 lifecycle worker: skip entire bucket when no lifecycle config is set 2023-08-30 14:38:19 +02:00
2996dc875f lifecycle worker: implement main functionality 2023-08-30 14:29:03 +02:00
a2e0e34db5 lifecycle: skeleton for lifecycle worker 2023-08-30 12:41:11 +02:00
f7b409f114 use a NaiveDate in data model, it serializes to string (iso 8601 format) 2023-08-30 11:24:01 +02:00
abf011c290 lifecycle: implement validation into garage's internal data structure 2023-08-29 18:22:03 +02:00
8041d9a827 s3: add xml structures to serialize/deserialize lifecycle configs 2023-08-29 17:44:17 +02:00
0b83e0558e bucket_table: data model for lifecycle configuration 2023-08-29 17:00:41 +02:00
2e90e1c124 Merge branch 'main' into next 2023-08-29 11:32:42 +02:00
32e5686ad8 Merge pull request 'Garage v0.8.3' (#619) from next-0.8 into main
Reviewed-on: Deuxfleurs/garage#619
2023-08-29 08:55:46 +00:00
06369c8f4a add garage_db dependency in garage_rpc 2023-08-28 17:08:21 +02:00
cece1be1bb bump version to 0.8.3 2023-08-28 13:17:26 +02:00
769b6fe054 fix test_website_check_domain 2023-08-28 12:40:28 +02:00
e66c78d6ea integration test: move json_body to root of crate 2023-08-28 12:32:57 +02:00
51011e68b1 move alpine linux info to binary package page 2023-08-28 12:20:34 +02:00
a54a1f5616 Merge pull request 'doc: Add information about Alpine Linux package to Quick Start' (#564) from jirutka/garage:alpine into next-0.8
Reviewed-on: Deuxfleurs/garage#564
2023-08-28 10:18:33 +00:00
9b4ce4a8ad admin api: refactor caddy check api code 2023-08-28 12:17:10 +02:00
2bbe2da5ad Merge pull request 'support index on path missing a trailing slash' (#612) from compat/index-without-trailing-slash into main
Reviewed-on: Deuxfleurs/garage#612
2023-08-28 10:15:01 +00:00
29353adbe5 Merge pull request 'cargo: Bump dependencies' (#606) from jpds/garage:cargo-bumps-230801 into main
Reviewed-on: Deuxfleurs/garage#606
2023-08-28 10:13:39 +00:00
c5cafa0000 web_server.rs: handle error properly and refactor 2023-08-28 12:05:14 +02:00
74478443ec update cargo.nix 2023-08-28 11:31:40 +02:00
Jonathan Davies
d66d81ae2d cargo: Updated gethostname v0.2.3 -> v0.4.3. 2023-08-28 09:30:27 +00:00
Jonathan Davies
7d8296ec59 cargo: Updated pretty_env_logger v0.4.0 -> v0.5.0. 2023-08-28 09:30:27 +00:00
Jonathan Davies
f607ac6792 garage/api: cargo: Updated idna dependency to 0.4. 2023-08-28 09:30:27 +00:00
Jonathan Davies
96d1d81ab7 garage/db: cargo: Updated rusqlite to 0.29. 2023-08-28 09:30:27 +00:00
Jonathan Davies
5185701aa8 cargo: Updated:
* addr2line v0.19.0 -> v0.20.0
 * async-compression v0.4.0 -> v0.4.1
 * clap v4.3.8 -> v4.3.19
 * hyper v0.14.26 -> v0.14.27
 * ipnet v2.7.2 -> v2.8.0
 * rmp v0.8.11 -> v0.8.12
 * serde v1.0.164 -> v1.0.188
 * tokio v1.29.0 -> v1.31.0
 * zstd v0.12.3+zstd.1.5.2 -> v0.12.4
 * Others in `cargo update`
2023-08-28 09:30:27 +00:00
d539a56d3a Merge pull request 'Support {s3,web}.root_domains for the Caddy on-demand TLS endpoint (<admin>/check?domain=xx)' (#610) from bug/support-root-domains-on-demand-tls into main
Reviewed-on: Deuxfleurs/garage#610
2023-08-28 09:18:13 +00:00
bd50333ade Merge pull request 'reverse-proxy.md: Added caching section for Caddy.' (#614) from jpds/garage:caddy-cache into main
Reviewed-on: Deuxfleurs/garage#614
2023-08-28 08:51:33 +00:00
170c6a2eac Merge pull request 'backup.md: Added restic-android note.' (#616) from jpds/garage:doc-restic-android into main
Reviewed-on: Deuxfleurs/garage#616
2023-08-28 08:50:57 +00:00
47e7f9e122 another typo 2023-08-19 20:29:24 +00:00
5ffcdb4634 fix typo 2023-08-19 15:17:51 +00:00
Jonathan Davies
7f7d85654d backup.md: Added restic-android note. 2023-08-18 18:02:19 +01:00
Jonathan Davies
245a0882e1 reverse-proxy.md: Added caching section for Caddy. 2023-08-16 11:49:52 +01:00
63da1d2443
support index on path missing a trailing slash 2023-08-08 15:28:57 +02:00
24e533f262
support {s3,web}.root_domains in /check endpoint 2023-08-08 11:05:42 +02:00
67b1457c77 Merge pull request 'post_object.rs: Fixed typos / grammar.' (#607) from jpds/garage:post-objects-typos into main
Reviewed-on: Deuxfleurs/garage#607
2023-08-04 07:09:21 +00:00
Jonathan Davies
59bfc68f2e post_object.rs: Fixed typos / grammar. 2023-08-01 15:31:39 +01:00
a98855157b Merge pull request 'operations/durability-repairs-md: Fix typo' (#604) from maxjustus/garage:main into main
Reviewed-on: Deuxfleurs/garage#604
2023-07-28 14:31:51 +00:00
4d7bbf7878 operations/durability-repairs-md: Fix typo 2023-07-24 10:01:48 -07:00
18eb73d52e Merge pull request 'flake-compat: use nix-community fork' (#599) from flokli/garage:flake-compat into main
Reviewed-on: Deuxfleurs/garage#599
2023-07-18 21:54:51 +00:00
79ca8e76a4 nix/common.nix: use pattern from nix-community/flake-compat
This is still a bit confusing, as normally the flake.defaultNix attrset
gets exposed via a top-level default.nix, but at least it brings us
closer to that.
2023-07-16 12:52:14 +03:00
1bbf604224 flake.nix: switch to nix-community/flake-compat
edolstra/flake-compat is unmaintained.

cargo2nix also still pulls in edolstra/flake-compat, make it follow the
nix-community one.
2023-07-16 12:40:47 +03:00
6ba611361e Merge pull request 'tree-wide: fix some typos' (#598) from flokli/garage:fix-typos into main
Reviewed-on: Deuxfleurs/garage#598
2023-07-14 15:51:44 +00:00
c855284760 src/util: fix typo 2023-07-14 14:25:40 +03:00
b1ca1784a1 src/garage/cli: fix typo 2023-07-14 14:25:33 +03:00
f0b7a0af3d doc/drafts: fix typo 2023-07-14 14:25:14 +03:00
194549ca46 doc/book: fix typo 2023-07-14 14:24:40 +03:00
202d3f0e3c doc/api: fix typo 2023-07-14 14:24:27 +03:00
7605d0cb11 Merge pull request 'cargo: tokio-1.29 and async-compression-0.4' (#593) from jpds/garage:tokio-1.29 into main
Reviewed-on: Deuxfleurs/garage#593
2023-07-03 17:03:09 +00:00
031804171a Update Cargo.nix 2023-07-03 11:33:36 +02:00
Jonathan Davies
aee0d97f22 cargo: Updated async-compression to 0.4. 2023-06-28 11:17:16 +01:00
Jonathan Davies
098c388f1b cargo: Updated tokio to 1.29. 2023-06-28 11:16:41 +01:00
e716320b0a Merge pull request 'cargo: roxmltree-0.18 and aws-sdk-s3-0.28 bump' (#591) from jpds/garage:roxmltree-0.18 into main
Reviewed-on: Deuxfleurs/garage#591
2023-06-27 17:20:58 +00:00
e466edbaec Merge pull request 'introduce dedicated return type for PollRange' (#590) from trinity-1686a/garage:k2v-client-poll-range-result into main
Reviewed-on: Deuxfleurs/garage#590
2023-06-27 08:28:26 +00:00
76355453dd Update Cargo.nix 2023-06-27 10:23:02 +02:00
ee494f5aa2 Merge pull request 'don't build sqlite by default' (#592) from trinity-1686a/garage:dont-build-sqlite into main
Reviewed-on: Deuxfleurs/garage#592
2023-06-27 08:14:38 +00:00
Jonathan Davies
f31d98097a Cargo.lock: Updated. 2023-06-26 18:03:47 +01:00
Jonathan Davies
a6da7e588f tests/bucket.rs: Adjusted as previously used function is now private. 2023-06-26 18:03:43 +01:00
e5835704b7 don't build sqlite by default
`bundled-libs` is enabled by default, and causes sqlite to be built too,
even if the sqlite backend isn't enabled.
2023-06-26 11:15:11 +02:00
Jonathan Davies
7f8bf2d801 src/garage/tests: Updated types for aws-sdk-s3 bump. 2023-06-25 21:31:35 +01:00
Jonathan Davies
4297233d3e garage/Cargo.toml: Updated aws-sdk-s3 to 0.28, added aws-config. 2023-06-25 21:17:15 +01:00
Jonathan Davies
b94ba47f29 api/Cargo.toml: Updated roxmltree to 0.18. 2023-06-24 14:15:26 +01:00
33b3cf8e22 introduce dedicated return type for PollRange 2023-06-24 10:17:20 +02:00
736083063f Merge pull request 'doc: Added ejabberd S3 section' (#588) from jpds/garage:doc-ejabberd-s3 into main
Reviewed-on: Deuxfleurs/garage#588
2023-06-20 09:23:43 +00:00
Jonathan Davies
a5ae566e0b apps/index.md: Fixed endpoint URL example. 2023-06-19 10:15:30 +01:00
Jonathan Davies
185f9e78f3 operations/durability-repairs.md: Added note about randomized scrub times. 2023-06-19 10:15:30 +01:00
Jonathan Davies
fb971a5f01 cookbook/encryption.md: Added Cyberduck note. 2023-06-19 10:15:30 +01:00
Jonathan Davies
6af2cde23f cookbook/encryption.md: Added note on XMPP. 2023-06-19 10:15:30 +01:00
Jonathan Davies
97eb389274 docs/apps: Added ejabberd section. 2023-06-19 10:15:30 +01:00
8ef42c9609 admin docs: reformatting, key admin: add check 2023-06-14 17:19:25 +02:00
a83a092c03 admin: uniformize layout api and improve code 2023-06-14 17:12:37 +02:00
7895f99d3a admin and cli: hide secret keys unless asked 2023-06-14 16:56:15 +02:00
4a82f6380e admin api: move all endpoints to v1/ by default (v0/ still supported) 2023-06-14 14:15:51 +02:00
28cc9f178a admin api: make name optionnal for CreateKey 2023-06-14 13:56:37 +02:00
2c83006608 admin api: fix doc in drafts 2023-06-14 13:54:34 +02:00
35c108b85d admin api: switch GetClusterHealth to camelcase (fix #381 again) 2023-06-14 13:53:19 +02:00
52376d47ca admin api: change cluster status/layout to use lists and not maps (fix #377) 2023-06-14 13:45:27 +02:00
187240e539 Merge branch 'main' into next 2023-06-14 13:02:46 +02:00
5e291c64b3 Merge pull request 'Documentation updates' (#587) from doc-updates into main
Reviewed-on: Deuxfleurs/garage#587
2023-06-14 10:57:32 +00:00
9092c71a01 doc: encryption organization 2023-06-14 12:51:47 +02:00
120f8b3bfb doc: better doc on systemd's DynamicUser (fix #430) 2023-06-14 12:39:46 +02:00
39c3738a07 Add a page about encryption (fix #416) 2023-06-14 12:39:46 +02:00
7169ee6ee6 doc: reformulate in monitoring page 2023-06-14 12:39:46 +02:00
dd7533a260 doc: add an operations&maintenance section and move some pages there 2023-06-14 12:39:40 +02:00
9233661967 Add documentation on durability and repair procedures (fix #219) 2023-06-14 11:54:21 +02:00
3aadba724d doc: english improvement 2023-06-14 11:21:56 +02:00
5a186be363 Doc: update goals, add docker alias
Fix #235
2023-06-14 11:09:31 +02:00
5670367126 multipartupload in test: add forgotten timestamp 2023-06-13 23:10:46 +02:00
cda957b4b1 update netapp's rmp-serde dependency to v1.1 2023-06-13 17:34:49 +02:00
90b2d43eb4 Merge branch 'main' into next 2023-06-13 17:14:11 +02:00
01346143ca Merge pull request 'Split src/garage/admin.rs into smaller files' (#586) from main-split-admin into main
Reviewed-on: Deuxfleurs/garage#586
2023-06-13 14:56:34 +00:00
eb9cecf05c Split garage/admin.rs into smaller files 2023-06-13 16:46:28 +02:00
802ed75721 move admin.rs to admin/mod.rs, before splitting 2023-06-13 16:42:14 +02:00
bf19a44fd9 admin API: add missing camelCase conversions (fix #381) 2023-06-13 16:15:50 +02:00
7126f3e1d1 garage key import: add checks and --yes CLI flag (fix #278) 2023-06-13 15:56:48 +02:00
fc29548933 Merge pull request 'fix timestamps wrapping around in garage block list-errors (fix #584)' (#585) from fix-future-timestamps into main
Reviewed-on: Deuxfleurs/garage#585
2023-06-13 12:51:16 +00:00
942c1f1bfe multipart uploads: save timestamp 2023-06-13 10:48:22 +02:00
1ea4937c8b fix timestamps wrapping around in garage block list-errors (fix #584) 2023-06-12 20:07:33 +02:00
0a06fda0da Merge pull request 'Fix #204 (full Multipart Uploads semantics)' (#553) from nlnet-task1 into next
Reviewed-on: Deuxfleurs/garage#553
2023-06-09 15:34:09 +00:00
3d477906d4 properly delete multipart uploads after completion 2023-06-09 17:13:27 +02:00
e645bbd3ce smoke test: add multipart upload test with part re-upload 2023-06-09 16:23:37 +02:00
58563ed700 Add multipart upload using aws s3api 2023-06-09 16:23:37 +02:00
a6cc563bdd UploadPart: automatic cleanup of version (and reference blocked) when interrupted 2023-06-09 16:23:37 +02:00
c14d3735e5 Add test for multipart uploads and fix part renumbering 2023-06-09 16:23:37 +02:00
53bf2f070c undo sort_key() returning Cow 2023-06-09 16:23:37 +02:00
412ab77b08 comments and clippy lint fixes 2023-06-09 16:23:37 +02:00
511e07ecd4 fix mpu counter (add missing workers) and report info at appropriate places 2023-06-09 16:23:37 +02:00
4ea53dc759 Add multipart upload repair 2023-06-09 16:23:37 +02:00
058518c22b refactor repair workers with a trait 2023-06-09 16:23:37 +02:00
8644376ac2 fix test; simplify code 2023-06-09 16:23:37 +02:00
7ad7dae5d4 fix s3 list test 2023-06-09 16:23:37 +02:00
75a0e01372 fix online repair 2023-06-09 16:23:37 +02:00
bb176ebcb8 cargo fmt 2023-06-09 16:23:37 +02:00
c1e1764f17 move git-version dependency to main crate to reduce rebuilds 2023-06-09 16:23:37 +02:00
87be8eeb93 updaet block admin for new multipartupload models 2023-06-09 16:23:37 +02:00
82e75c0e29 Adapt S3 API code to use new multipart upload models
- Create and PutPart
- completemultipartupload
- upload part copy
- list_parts
2023-06-09 16:23:37 +02:00
38d6ac4295 New multipart upload table layout 2023-06-09 16:23:37 +02:00
6005491cd8 Use Cow<[u8]> for sort keys 2023-06-09 16:23:37 +02:00
ea3bfd2ab1 Minio tests for multipart upload behaviour:
- upload part renumbering test
- part skipping test
2023-06-09 16:23:37 +02:00
e7e164a280 Make fsync an option for meta and data 2023-06-09 16:23:21 +02:00
1e466b11eb Revert integration tests to using Sled as LMDB causes failures 2023-06-09 13:23:08 +02:00
865f0c7d0c Add LMDB to debug builds 2023-06-09 12:04:28 +02:00
906fe78b24 Integration tests: print logs when fails 2023-06-09 12:03:44 +02:00
6aec73b641 Merge pull request 'payload.rs: Fixed two typoes' (#581) from jpds/garage:payload-typoes into main
Reviewed-on: Deuxfleurs/garage#581
2023-06-09 08:59:47 +00:00
Jonathan Davies
8a945ee996 payload.rs: Surround / in inverted commas. 2023-06-06 16:26:06 +01:00
Jonathan Davies
180992d0f1 payload.rs: Fixed typo in error message. 2023-06-06 16:25:29 +01:00
8a74e1c2bd Split garage/admin.rs into smaller files 2023-06-06 15:39:15 +02:00
44548a9114 Merge pull request 'feature: Register consul services with agent API' (#567) from unrob/garage:roberto/consul-agent-registration into main
Reviewed-on: Deuxfleurs/garage#567
Reviewed-by: Alex <alex@adnab.me>
2023-06-02 14:35:00 +00:00
Roberto Hidalgo
32ad4538ee fix references to old config names 2023-05-22 08:57:15 -06:00
Roberto Hidalgo
ef8a7add08 set default for [consul-services] api 2023-05-22 08:57:15 -06:00
Roberto Hidalgo
2d46d24d06 update docs 2023-05-22 08:57:15 -06:00
Roberto Hidalgo
b770504126 simplify code according to feedback 2023-05-22 08:57:15 -06:00
Roberto Hidalgo
6b69404f1a rename mode to consul_http_api 2023-05-22 08:57:15 -06:00
Roberto Hidalgo
011f473048 revert rpc/Cargo.toml 2023-05-22 08:57:15 -06:00
Roberto Hidalgo
fd7dbea5b8 follow feedback, fold into existing feature 2023-05-22 08:57:15 -06:00
Roberto Hidalgo
bd6485565e allow additional ServiceMeta, docs 2023-05-22 08:57:15 -06:00
Roberto Hidalgo
4d6e6fc155 cargo fmt 2023-05-22 08:57:15 -06:00
Roberto Hidalgo
02ba9016ab register consul services against local agent instead of catalog api 2023-05-22 08:57:15 -06:00
9d833bb7ef Merge pull request 'K2V-client improvements' (#577) from k2v-client-aws-sigv4 into main
Reviewed-on: Deuxfleurs/garage#577
2023-05-22 09:03:08 +00:00
c3d3b837eb bump k2v-client to v0.0.4 2023-05-22 10:47:15 +02:00
130e01505b Fix k2v_client with unicode in partition keys 2023-05-22 10:45:09 +02:00
e2ce5970c6 Add basic k2v_client integration tests 2023-05-22 10:45:06 +02:00
644e872264 Port k2v-client to aws-sigv4 since rusoto_signature is deprecated 2023-05-19 12:08:29 +02:00
03efc191c1 Merge pull request 'K2V: double urlencoding' (#574) from fix-k2v-urlencoding into main
Reviewed-on: Deuxfleurs/garage#574
2023-05-18 09:33:03 +00:00
4420db7310 add tracing to k2v-client 2023-05-18 11:18:21 +02:00
746b0090e4 k2v signature verification: double urlencoding (see comment in source code) 2023-05-18 11:18:06 +02:00
c26a4308b4 Merge pull request 'Split format_table into separate crate and reduce k2v-client dependencies' (#572) from split-format-table into main
Reviewed-on: Deuxfleurs/garage#572
2023-05-17 12:33:45 +00:00
19639705e6 Mark sled as deprecated, make lmdb default, and improve sqlite and lmdb defaults 2023-05-17 14:30:53 +02:00
217d429937 fix clippy lint in format-table crate 2023-05-17 13:06:37 +02:00
a1cec2cd60 Split format_table into separate crate and reduce k2v-client dependencies 2023-05-17 13:01:37 +02:00
b66f247580 Merge pull request 'fixes to K2V client' (#571) from k2v-client-fixes into main
Reviewed-on: Deuxfleurs/garage#571
2023-05-16 20:20:31 +00:00
16f2a32bb7 cargo fmt 2023-05-16 19:46:57 +02:00
472444ed8e k2v-client 0.0.2 2023-05-16 19:46:57 +02:00
bb03805b58 k2v-cli: fix sort_key being partition_key and specify which key 2023-05-16 19:46:57 +02:00
e4f955d672 fix base64 uses 2023-05-16 19:46:56 +02:00
ea9b15f669 Merge pull request 'cargo: tokio-1.28 and hyper-0.14.26 update' (#569) from jpds/garage:tokio-1.28 into main
Reviewed-on: Deuxfleurs/garage#569
2023-05-11 10:16:33 +00:00
2e6bb3f766 update Cargo.nix 2023-05-11 11:34:18 +02:00
375270afd1 Merge pull request '*: apply clippy recommendations.' (#570) from jpds/garage:clippy-fixes into main
Reviewed-on: Deuxfleurs/garage#570
2023-05-11 09:33:03 +00:00
Jonathan Davies
c783194e8b *: apply clippy recommendations. 2023-05-09 20:49:34 +01:00
Jonathan Davies
fdcd7dee5a Cargo.lock: Updated for:
* tokio 1.28
 * hyper 0.14.26
2023-05-09 14:43:52 +01:00
Jonathan Davies
0f0795103d block/Cargo.toml: Bump tokio-util to 0.7. 2023-05-09 14:33:21 +01:00
Jonathan Davies
c9d26e8c50 k2v-client/Cargo.toml: Make tokio dep match other packages. 2023-05-09 14:33:00 +01:00
351d734e6c Merge branch 'main' into next 2023-05-09 12:40:08 +02:00
b925f53dc3 Merge pull request 'move git-version dependency to main crate to reduce rebuilds' (#568) from move-git-version into main
Reviewed-on: Deuxfleurs/garage#568
2023-05-09 09:53:33 +00:00
2f495575d8 Merge pull request 'block/manager.rs: Prioritize raw blocks when no compression configured' (#566) from jpds/garage:skip-compressed-blocks-scrub-no-compression into main
Reviewed-on: Deuxfleurs/garage#566
2023-05-09 09:39:48 +00:00
9e0a9c1c15 move git-version dependency to main crate to reduce rebuilds 2023-05-09 11:35:32 +02:00
Jonathan Davies
9c788059e2 block/manager.rs: In is_block_compressed - check which compression_level
is configured on a node and check for raw block first if compression is
disabled (to help reduce syscalls during a scrub).
2023-05-09 10:28:19 +01:00
5684e1990c Merge pull request 'Really allow to disable sled feature' (#563) from jirutka/garage:workspace-deps into main
Reviewed-on: Deuxfleurs/garage#563
2023-05-09 09:08:35 +00:00
14c50f2f84 Merge pull request 'Fix undefined macro warn! on 32-bit' (#562) from jirutka/garage:fix-undefined-warn into main
Reviewed-on: Deuxfleurs/garage#562
2023-05-09 08:52:11 +00:00
0fab9c3b8c Merge pull request 'Helm: Include newer config parameters as values' (#565) from jonatan/garage:main into main
Reviewed-on: Deuxfleurs/garage#565
2023-05-09 08:49:00 +00:00
75759a163c Allow to really disable sled feature 2023-05-09 08:46:15 +00:00
d2deee0b8b Declare garage crates using workspace.dependencies
This will allow to really disable "sled" feature without declaring
`default-features = false` in every Cargo.toml where garage_db and
garage_model is used.

See https://doc.rust-lang.org/cargo/reference/workspaces.html#the-dependencies-table
2023-05-09 08:46:15 +00:00
8499cd5c21 Merge pull request 'Remove unnecessary/unused "timeago" features' (#559) from jirutka/garage:timeago-features into main
Reviewed-on: Deuxfleurs/garage#559
2023-05-09 08:44:22 +00:00
4ea7983093
Helm: Increment patch version 2023-05-08 08:03:21 +02:00
d5e39d11eb
Helm: Include newer config parameters as values
Add all missing parameters from the reference manual.
Primarily to enable the use of the new lmdb engine
2023-05-08 07:47:31 +02:00
06caa12d49 doc: Add information about Alpine Linux package to Quick Start 2023-05-07 19:28:43 +02:00
6d3ace1ea9 Fix undefined macro warn! on 32-bit
Compiling garage_db v0.8.2 (garage-0.8.2/src/db)
    error: cannot find macro `warn` in this scope
       --> src/db/lmdb_adapter.rs:352:2
        |
    352 |     warn!("LMDB is not recommended on 32-bit systems, database size will be limited");
        |     ^^^^
        |
        = help: consider importing this macro:
                tracing::warn
        = note: `warn` is in scope, but it is an attribute: `#[warn]`
    error: could not compile `garage_db` due to previous error
2023-05-07 17:01:44 +02:00
833cf082da Remove unnecessary/unused "timeago" features
To decrease dependency bloat and binary size.
2023-05-07 01:03:54 +02:00
a1fcf1b175 Merge branch 'main' into next 2023-04-25 16:58:57 +02:00
1ecd88c01f Merge pull request 'Update rust toolchain to 1.68 and simplify Nix stuff' (#554) from nix-update-simplify into main
Reviewed-on: Deuxfleurs/garage#554
2023-04-25 14:56:49 +00:00
5efcdc0de3 Update rust toolchain to 1.68 and simplify Nix stuff 2023-04-25 14:46:47 +02:00
fa78d806e3 Merge branch 'main' into next 2023-04-25 12:34:26 +02:00
a16eb7e4b8 Merge pull request 'api/Cargo.toml: Bumped quick-xml to version 0.26.' (#552) from jpds/garage:quick-xml-0.26 into main
Reviewed-on: Deuxfleurs/garage#552
2023-04-24 09:00:00 +00:00
6742070517 Merge pull request 'block/repair.rs: Added log entries for scrub start/finish.' (#551) from jpds/garage:scrub-log into main
Reviewed-on: Deuxfleurs/garage#551
2023-04-24 08:29:36 +00:00
6894878146 update cargo.nix 2023-04-24 10:26:14 +02:00
02b0ba5f44 Merge pull request 'cookbook/real-world: fix typo' (#549) from yuka/garage:main into main
Reviewed-on: Deuxfleurs/garage#549
2023-04-24 08:24:55 +00:00
Jonathan Davies
fb3bd11dce block/repair.rs: Added log entries for scrub start/finish. 2023-04-23 22:22:26 +01:00
Jonathan Davies
c168383113 api/Cargo.toml: Bumped quick-xml to version 0.26. 2023-04-23 20:14:28 +01:00
04a0063df9 cookbook/real-world: fix typo 2023-04-21 16:46:58 +00:00
654999e254 Update Cargo.nix 2023-01-26 15:50:54 +01:00
Jonathan Davies
db56d4658f util/Cargo.toml: Updated rmp-serde from 0.15 to 1.1. 2023-01-26 11:03:43 +00:00
12a4e1f303
Merge branch 'optimal-layout' into next 2023-01-11 17:50:42 +01:00
84b4a868e3
Migration of cluster layout from v0.8 to v0.9 2023-01-11 17:47:46 +01:00
4f409f73dc Merge pull request 'Changed all instances of assignation to assignment' (#465) from jpds/garage:assignments-correction into next
Reviewed-on: Deuxfleurs/garage#465
2023-01-11 16:05:27 +00:00
597d64b31a change in gitignore 2023-01-09 16:06:47 +01:00
e3cc7a89b0 First draft of t a preprint describing the layout computation algorithm 2023-01-09 16:05:20 +01:00
Jonathan Davies
cb07e6145c Changed all instances of assignation to assignment. 2023-01-05 11:09:25 +00:00
80e4abb98d Merge pull request 'Changed all instances of 'key new' to 'key create' to make it the same as the bucket commands.' (#459) from jpds/garage:key-create-standardize into next
Reviewed-on: Deuxfleurs/garage#459
2023-01-04 10:35:49 +00:00
570e5e5bbb
Merge branch 'main' into next 2023-01-04 11:34:43 +01:00
Jonathan Davies
8be862aa19 Changed all instances of 'key new' to 'key create' to make it consistent as bucket commands issued normally around the same time. 2023-01-03 11:11:12 +00:00
6e44369cbc Merge pull request 'Optimal layout assignation algorithm' (#296) from optimal-layout into next
Reviewed-on: Deuxfleurs/garage#296
2022-12-11 17:41:53 +00:00
2c2e65ad8b
Merge commit 'ec12d6c' into next 2022-12-11 18:41:15 +01:00
9d83364ad9
itertools .unique() doesn't require sorted items 2022-12-11 18:30:02 +01:00
ec12d6c8dd
Slightly simplify code at places 2022-11-08 16:15:45 +01:00
217abdca18
Fix HTTP return code 2022-11-08 15:38:53 +01:00
fc2729cd81
Fix integration test 2022-11-08 15:19:46 +01:00
d75b37b018
Return more info when layout's .check() fails, fix compilation, fix test 2022-11-08 14:58:39 +01:00
73a4ca8b15
Use bytes as capacity units 2022-11-07 21:12:11 +01:00
fd5bc142b5
Ensure .sort() is called before counting unique items 2022-11-07 20:29:25 +01:00
ea5afc2511
Style improvements 2022-11-07 20:11:30 +01:00
28d7a49f63
Merge branch 'main' into optimal-layout 2022-11-07 12:20:59 +01:00
3039bb5d43
rm .gitattributes 2022-10-13 12:40:42 +02:00
bcdd1e0c33 Added some comment 2022-10-11 18:29:21 +02:00
e5664c9822 Improved the statistics displayed in layout show
corrected a few bugs
2022-10-11 17:17:13 +02:00
4abab246f1 cargo fmt 2022-10-10 17:21:13 +02:00
fcf9ac674a Tests written in layout.rs
added staged_parameters to ClusterLayout
removed the serde(default) -> will need a migration function
2022-10-10 17:19:25 +02:00
911eb17bd9 corrected warnings of cargo clippy 2022-10-06 14:53:57 +02:00
9407df60cc Corrected two bugs:
- self.node_id_vec was not properly updated when the previous ring was empty
- ClusterLayout::merge was not considering changes in the layout parameters
2022-10-06 12:54:51 +02:00
a951b6c452 Added a CLI command to update the parameters for the layout computation (for now, only the zone redundancy) 2022-10-05 16:04:19 +02:00
ceac3713d6 modifications in several files to :
- have consistent error return types
- store the zone redundancy in a Lww
- print the error and message in the CLI (TODO: for the server Api, should msg be returned in the body response?)
2022-10-05 15:29:48 +02:00
829f815a89 Merge remote-tracking branch 'origin/main' into optimal-layout 2022-10-04 18:14:49 +02:00
99f96b9564 deleted zone_redundancy from System struct 2022-10-04 18:09:24 +02:00
bd842e1388 Correction of a few bugs in the tests, modification of ClusterLayout::check 2022-09-22 19:30:01 +02:00
7f3249a237 New version of the algorithm that calculate the layout.
It takes as paramters the replication factor and the zone redundancy, computes the
largest partition size reachable with these constraints, and among the possible
assignation with this partition size, it computes the one that moves the least number
of partitions compared to the previous assignation.
This computation uses graph algorithms defined in graph_algo.rs
2022-09-21 14:39:59 +02:00
c4adbeed51 Added the section with description proofs of the parametric assignment computation in the optimal layout report 2022-09-10 13:51:12 +02:00
d38fb6c250 ignore log files in commit 2022-09-08 12:43:33 +02:00
81083dd415 Added a first draft version of the algorithm and analysis for the non-strict mode. 2022-08-19 21:21:41 +02:00
7b2c065c82 Merge branch 'optimal-layout' of https://git.deuxfleurs.fr/Deuxfleurs/garage into optimal-layout 2022-07-19 13:30:49 +02:00
03e3a1bd15 Added the latex report on the optimal layout algorithm 2022-07-18 22:35:29 +02:00
617f28bfa4
Correct small formatting issue 2022-05-05 14:21:57 +02:00
948ff93cf1 Corrected the warnings and errors issued by cargo clippy 2022-05-01 16:05:39 +02:00
3ba2c5b424
updated cargo.lock 2022-05-01 10:11:43 +02:00
2aeaddd5e2
Apply cargo fmt 2022-05-01 09:57:05 +02:00
c1d1646c4d
Change the way new layout assignations are computed.
The function now computes an optimal assignation (with respect to partition size) that minimizes the distance to the former assignation, using flow algorithms.

This commit was written by Mendes Oulamara <mendes.oulamara@pm.me>
2022-05-01 09:54:19 +02:00
460 changed files with 85755 additions and 14422 deletions

3
.cargo/config.toml Normal file
View file

@ -0,0 +1,3 @@
[target.x86_64-unknown-linux-gnu]
linker = "clang"
rustflags = ["-C", "link-arg=-fuse-ld=mold"]

View file

@ -1,288 +0,0 @@
---
kind: pipeline
name: default
node:
nix-daemon: 1
steps:
- name: check formatting
image: nixpkgs/nix:nixos-22.05
commands:
- nix-shell --attr rust --run "cargo fmt -- --check"
- name: build
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr clippy.amd64 --argstr git_version ${DRONE_TAG:-$DRONE_COMMIT}
- name: unit + func tests
image: nixpkgs/nix:nixos-22.05
environment:
GARAGE_TEST_INTEGRATION_EXE: result-bin/bin/garage
commands:
- nix-build --no-build-output --attr clippy.amd64 --argstr git_version ${DRONE_TAG:-$DRONE_COMMIT}
- nix-build --no-build-output --attr test.amd64
- ./result/bin/garage_db-*
- ./result/bin/garage_api-*
- ./result/bin/garage_model-*
- ./result/bin/garage_rpc-*
- ./result/bin/garage_table-*
- ./result/bin/garage_util-*
- ./result/bin/garage_web-*
- ./result/bin/garage-*
- ./result/bin/integration-*
- rm result
- name: integration tests
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr clippy.amd64 --argstr git_version ${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr integration --run ./script/test-smoke.sh || (cat /tmp/garage.log; false)
trigger:
event:
- custom
- push
- pull_request
- tag
- cron
---
kind: pipeline
type: docker
name: release-linux-amd64
node:
nix-daemon: 1
steps:
- name: build
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr pkgs.amd64.release --argstr git_version ${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr rust --run "./script/not-dynamic.sh result-bin/bin/garage"
- name: integration
image: nixpkgs/nix:nixos-22.05
commands:
- nix-shell --attr integration --run ./script/test-smoke.sh || (cat /tmp/garage.log; false)
- name: push static binary
image: nixpkgs/nix:nixos-22.05
environment:
AWS_ACCESS_KEY_ID:
from_secret: garagehq_aws_access_key_id
AWS_SECRET_ACCESS_KEY:
from_secret: garagehq_aws_secret_access_key
TARGET: "x86_64-unknown-linux-musl"
commands:
- nix-shell --attr release --run "to_s3"
- name: docker build and publish
image: nixpkgs/nix:nixos-22.05
environment:
DOCKER_AUTH:
from_secret: docker_auth
DOCKER_PLATFORM: "linux/amd64"
CONTAINER_NAME: "dxflrs/amd64_garage"
HOME: "/kaniko"
commands:
- mkdir -p /kaniko/.docker
- echo $DOCKER_AUTH > /kaniko/.docker/config.json
- export CONTAINER_TAG=${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr release --run "to_docker"
trigger:
event:
- promote
- cron
---
kind: pipeline
type: docker
name: release-linux-i386
node:
nix-daemon: 1
steps:
- name: build
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr pkgs.i386.release --argstr git_version ${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr rust --run "./script/not-dynamic.sh result-bin/bin/garage"
- name: integration
image: nixpkgs/nix:nixos-22.05
commands:
- nix-shell --attr integration --run ./script/test-smoke.sh || (cat /tmp/garage.log; false)
- name: push static binary
image: nixpkgs/nix:nixos-22.05
environment:
AWS_ACCESS_KEY_ID:
from_secret: garagehq_aws_access_key_id
AWS_SECRET_ACCESS_KEY:
from_secret: garagehq_aws_secret_access_key
TARGET: "i686-unknown-linux-musl"
commands:
- nix-shell --attr release --run "to_s3"
- name: docker build and publish
image: nixpkgs/nix:nixos-22.05
environment:
DOCKER_AUTH:
from_secret: docker_auth
DOCKER_PLATFORM: "linux/386"
CONTAINER_NAME: "dxflrs/386_garage"
HOME: "/kaniko"
commands:
- mkdir -p /kaniko/.docker
- echo $DOCKER_AUTH > /kaniko/.docker/config.json
- export CONTAINER_TAG=${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr release --run "to_docker"
trigger:
event:
- promote
- cron
---
kind: pipeline
type: docker
name: release-linux-arm64
node:
nix-daemon: 1
steps:
- name: build
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr pkgs.arm64.release --argstr git_version ${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr rust --run "./script/not-dynamic.sh result-bin/bin/garage"
- name: push static binary
image: nixpkgs/nix:nixos-22.05
environment:
AWS_ACCESS_KEY_ID:
from_secret: garagehq_aws_access_key_id
AWS_SECRET_ACCESS_KEY:
from_secret: garagehq_aws_secret_access_key
TARGET: "aarch64-unknown-linux-musl"
commands:
- nix-shell --attr release --run "to_s3"
- name: docker build and publish
image: nixpkgs/nix:nixos-22.05
environment:
DOCKER_AUTH:
from_secret: docker_auth
DOCKER_PLATFORM: "linux/arm64"
CONTAINER_NAME: "dxflrs/arm64_garage"
HOME: "/kaniko"
commands:
- mkdir -p /kaniko/.docker
- echo $DOCKER_AUTH > /kaniko/.docker/config.json
- export CONTAINER_TAG=${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr release --run "to_docker"
trigger:
event:
- promote
- cron
---
kind: pipeline
type: docker
name: release-linux-arm
node:
nix-daemon: 1
steps:
- name: build
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr pkgs.arm.release --argstr git_version ${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr rust --run "./script/not-dynamic.sh result-bin/bin/garage"
- name: push static binary
image: nixpkgs/nix:nixos-22.05
environment:
AWS_ACCESS_KEY_ID:
from_secret: garagehq_aws_access_key_id
AWS_SECRET_ACCESS_KEY:
from_secret: garagehq_aws_secret_access_key
TARGET: "armv6l-unknown-linux-musleabihf"
commands:
- nix-shell --attr release --run "to_s3"
- name: docker build and publish
image: nixpkgs/nix:nixos-22.05
environment:
DOCKER_AUTH:
from_secret: docker_auth
DOCKER_PLATFORM: "linux/arm"
CONTAINER_NAME: "dxflrs/arm_garage"
HOME: "/kaniko"
commands:
- mkdir -p /kaniko/.docker
- echo $DOCKER_AUTH > /kaniko/.docker/config.json
- export CONTAINER_TAG=${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr release --run "to_docker"
trigger:
event:
- promote
- cron
---
kind: pipeline
type: docker
name: refresh-release-page
node:
nix-daemon: 1
steps:
- name: multiarch-docker
image: nixpkgs/nix:nixos-22.05
environment:
DOCKER_AUTH:
from_secret: docker_auth
HOME: "/root"
commands:
- mkdir -p /root/.docker
- echo $DOCKER_AUTH > /root/.docker/config.json
- export CONTAINER_TAG=${DRONE_TAG:-$DRONE_COMMIT}
- nix-shell --attr release --run "multiarch_docker"
- name: refresh-index
image: nixpkgs/nix:nixos-22.05
environment:
AWS_ACCESS_KEY_ID:
from_secret: garagehq_aws_access_key_id
AWS_SECRET_ACCESS_KEY:
from_secret: garagehq_aws_secret_access_key
commands:
- mkdir -p /etc/nix && cp nix/nix.conf /etc/nix/nix.conf
- nix-shell --attr release --run "refresh_index"
depends_on:
- release-linux-amd64
- release-linux-i386
- release-linux-arm64
- release-linux-arm
trigger:
event:
- promote
- cron
---
kind: signature
hmac: ac09a5a8c82502f67271f93afa1e1e21ce66383b8e24a6deb26b285cc1c378ba
...

47
.woodpecker/debug.yaml Normal file
View file

@ -0,0 +1,47 @@
when:
event:
- push
- tag
- pull_request
- deployment
- cron
- manual
steps:
- name: check formatting
image: nixpkgs/nix:nixos-22.05
commands:
- nix-shell --attr devShell --run "cargo fmt -- --check"
- name: build
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr clippy.amd64 --argstr git_version ${CI_COMMIT_TAG:-$CI_COMMIT_SHA}
- name: unit + func tests
image: nixpkgs/nix:nixos-22.05
environment:
GARAGE_TEST_INTEGRATION_EXE: result-bin/bin/garage
GARAGE_TEST_INTEGRATION_PATH: tmp-garage-integration
commands:
- nix-build --no-build-output --attr clippy.amd64 --argstr git_version ${CI_COMMIT_TAG:-$CI_COMMIT_SHA}
- nix-build --no-build-output --attr test.amd64
- ./result/bin/garage_db-*
- ./result/bin/garage_api-*
- ./result/bin/garage_model-*
- ./result/bin/garage_rpc-*
- ./result/bin/garage_table-*
- ./result/bin/garage_util-*
- ./result/bin/garage_web-*
- ./result/bin/garage-*
- GARAGE_TEST_INTEGRATION_DB_ENGINE=lmdb ./result/bin/integration-* || (cat tmp-garage-integration/stderr.log; false)
- nix-shell --attr ci --run "killall -9 garage" || true
- GARAGE_TEST_INTEGRATION_DB_ENGINE=sqlite ./result/bin/integration-* || (cat tmp-garage-integration/stderr.log; false)
- rm result
- rm -rv tmp-garage-integration
- name: integration tests
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr clippy.amd64 --argstr git_version ${CI_COMMIT_TAG:-$CI_COMMIT_SHA}
- nix-shell --attr ci --run ./script/test-smoke.sh || (cat /tmp/garage.log; false)

29
.woodpecker/publish.yaml Normal file
View file

@ -0,0 +1,29 @@
when:
event:
- deployment
- cron
depends_on:
- release
steps:
- name: refresh-index
image: nixpkgs/nix:nixos-22.05
secrets:
- source: garagehq_aws_access_key_id
target: AWS_ACCESS_KEY_ID
- source: garagehq_aws_secret_access_key
target: AWS_SECRET_ACCESS_KEY
commands:
- mkdir -p /etc/nix && cp nix/nix.conf /etc/nix/nix.conf
- nix-shell --attr ci --run "refresh_index"
- name: multiarch-docker
image: nixpkgs/nix:nixos-22.05
secrets:
- docker_auth
commands:
- mkdir -p /root/.docker
- echo $DOCKER_AUTH > /root/.docker/config.json
- export CONTAINER_TAG=${CI_COMMIT_TAG:-$CI_COMMIT_SHA}
- nix-shell --attr ci --run "multiarch_docker"

70
.woodpecker/release.yaml Normal file
View file

@ -0,0 +1,70 @@
when:
event:
- deployment
- cron
matrix:
include:
- ARCH: amd64
TARGET: x86_64-unknown-linux-musl
- ARCH: i386
TARGET: i686-unknown-linux-musl
- ARCH: arm64
TARGET: aarch64-unknown-linux-musl
- ARCH: arm
TARGET: armv6l-unknown-linux-musleabihf
steps:
- name: build
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr pkgs.${ARCH}.release --argstr git_version ${CI_COMMIT_TAG:-$CI_COMMIT_SHA}
- name: check is static binary
image: nixpkgs/nix:nixos-22.05
commands:
- nix-build --no-build-output --attr pkgs.${ARCH}.release --argstr git_version ${CI_COMMIT_TAG:-$CI_COMMIT_SHA}
- nix-shell --attr ci --run "./script/not-dynamic.sh result-bin/bin/garage"
- name: integration tests
image: nixpkgs/nix:nixos-22.05
commands:
- nix-shell --attr ci --run ./script/test-smoke.sh || (cat /tmp/garage.log; false)
when:
- matrix:
ARCH: amd64
- matrix:
ARCH: i386
- name: upgrade tests
image: nixpkgs/nix:nixos-22.05
commands:
- nix-shell --attr ci --run "./script/test-upgrade.sh v0.8.4 x86_64-unknown-linux-musl" || (cat /tmp/garage.log; false)
when:
- matrix:
ARCH: amd64
- name: push static binary
image: nixpkgs/nix:nixos-22.05
environment:
TARGET: "${TARGET}"
secrets:
- source: garagehq_aws_access_key_id
target: AWS_ACCESS_KEY_ID
- source: garagehq_aws_secret_access_key
target: AWS_SECRET_ACCESS_KEY
commands:
- nix-shell --attr ci --run "to_s3"
- name: docker build and publish
image: nixpkgs/nix:nixos-22.05
environment:
DOCKER_PLATFORM: "linux/${ARCH}"
CONTAINER_NAME: "dxflrs/${ARCH}_garage"
secrets:
- docker_auth
commands:
- mkdir -p /root/.docker
- echo $DOCKER_AUTH > /root/.docker/config.json
- export CONTAINER_TAG=${CI_COMMIT_TAG:-$CI_COMMIT_SHA}
- nix-shell --attr ci --run "to_docker"

3175
Cargo.lock generated

File diff suppressed because it is too large Load diff

6355
Cargo.nix

File diff suppressed because it is too large Load diff

View file

@ -3,6 +3,7 @@ resolver = "2"
members = [
"src/db",
"src/util",
"src/net",
"src/rpc",
"src/table",
"src/block",
@ -11,12 +12,141 @@ members = [
"src/web",
"src/garage",
"src/k2v-client",
"src/format-table",
]
default-members = ["src/garage"]
[workspace.dependencies]
# Internal Garage crates
format_table = { version = "0.1.1", path = "src/format-table" }
garage_api = { version = "1.0.1", path = "src/api" }
garage_block = { version = "1.0.1", path = "src/block" }
garage_db = { version = "1.0.1", path = "src/db", default-features = false }
garage_model = { version = "1.0.1", path = "src/model", default-features = false }
garage_net = { version = "1.0.1", path = "src/net" }
garage_rpc = { version = "1.0.1", path = "src/rpc" }
garage_table = { version = "1.0.1", path = "src/table" }
garage_util = { version = "1.0.1", path = "src/util" }
garage_web = { version = "1.0.1", path = "src/web" }
k2v-client = { version = "0.0.4", path = "src/k2v-client" }
# External crates from crates.io
arc-swap = "1.0"
argon2 = "0.5"
async-trait = "0.1.7"
backtrace = "0.3"
base64 = "0.21"
blake2 = "0.10"
bytes = "1.0"
bytesize = "1.1"
cfg-if = "1.0"
chrono = "0.4"
crc32fast = "1.4"
crc32c = "0.6"
crypto-common = "0.1"
digest = "0.10"
err-derive = "0.3"
gethostname = "0.4"
git-version = "0.3.4"
hex = "0.4"
hexdump = "0.1"
hmac = "0.12"
idna = "0.5"
itertools = "0.12"
ipnet = "2.9.0"
lazy_static = "1.4"
md-5 = "0.10"
mktemp = "0.5"
nix = { version = "0.27", default-features = false, features = ["fs"] }
nom = "7.1"
parse_duration = "2.1"
pin-project = "1.0.12"
pnet_datalink = "0.34"
rand = "0.8"
sha1 = "0.10"
sha2 = "0.10"
timeago = { version = "0.4", default-features = false }
xxhash-rust = { version = "0.8", default-features = false, features = ["xxh3"] }
aes-gcm = { version = "0.10", features = ["aes", "stream"] }
sodiumoxide = { version = "0.2.5-0", package = "kuska-sodiumoxide" }
kuska-handshake = { version = "0.2.0", features = ["default", "async_std"] }
clap = { version = "4.1", features = ["derive", "env"] }
pretty_env_logger = "0.5"
structopt = { version = "0.3", default-features = false }
syslog-tracing = "0.3"
tracing = "0.1"
tracing-subscriber = { version = "0.3", features = ["env-filter"] }
heed = { version = "0.11", default-features = false, features = ["lmdb"] }
rusqlite = "0.31.0"
r2d2 = "0.8"
r2d2_sqlite = "0.24"
async-compression = { version = "0.4", features = ["tokio", "zstd"] }
zstd = { version = "0.13", default-features = false }
quick-xml = { version = "0.26", features = [ "serialize" ] }
rmp-serde = "1.1.2"
serde = { version = "1.0", default-features = false, features = ["derive", "rc"] }
serde_bytes = "0.11"
serde_json = "1.0"
toml = { version = "0.8", default-features = false, features = ["parse"] }
# newer version requires rust edition 2021
k8s-openapi = { version = "0.21", features = ["v1_24"] }
kube = { version = "0.88", default-features = false, features = ["runtime", "derive", "client", "rustls-tls"] }
schemars = "0.8"
reqwest = { version = "0.11", default-features = false, features = ["rustls-tls-manual-roots", "json"] }
form_urlencoded = "1.0.0"
http = "1.0"
httpdate = "1.0"
http-range = "0.1"
http-body-util = "0.1"
hyper = { version = "1.0", default-features = false }
hyper-util = { version = "0.1", features = [ "full" ] }
multer = "3.0"
percent-encoding = "2.2"
roxmltree = "0.19"
url = "2.3"
futures = "0.3"
futures-util = "0.3"
tokio = { version = "1.0", default-features = false, features = ["net", "rt", "rt-multi-thread", "io-util", "net", "time", "macros", "sync", "signal", "fs"] }
tokio-util = { version = "0.7", features = ["compat", "io"] }
tokio-stream = { version = "0.1", features = ["net"] }
opentelemetry = { version = "0.17", features = [ "rt-tokio", "metrics", "trace" ] }
opentelemetry-prometheus = "0.10"
opentelemetry-otlp = "0.10"
opentelemetry-contrib = "0.9"
prometheus = "0.13"
# used by the k2v-client crate only
aws-sigv4 = { version = "1.1" }
hyper-rustls = { version = "0.26", features = ["http2"] }
log = "0.4"
thiserror = "1.0"
# ---- used only as build / dev dependencies ----
assert-json-diff = "2.0"
rustc_version = "0.4.0"
static_init = "1.0"
aws-config = "1.1.4"
aws-sdk-config = "1.13"
aws-sdk-s3 = "1.14"
[profile.dev]
#lto = "thin" # disabled for now, adds 2-4 min to each CI build
lto = "off"
[profile.release]
debug = true
lto = true
codegen-units = 1
opt-level = "s"
strip = true

View file

@ -1,4 +1,4 @@
Garage [![Build Status](https://drone.deuxfleurs.fr/api/badges/Deuxfleurs/garage/status.svg?ref=refs/heads/main)](https://drone.deuxfleurs.fr/Deuxfleurs/garage)
Garage [![status-badge](https://woodpecker.deuxfleurs.fr/api/badges/1/status.svg)](https://woodpecker.deuxfleurs.fr/repos/1)
===
<p align="center" style="text-align:center;">

View file

@ -40,7 +40,6 @@ in {
features = [
"garage/bundled-libs"
"garage/k2v"
"garage/sled"
"garage/lmdb"
"garage/sqlite"
];

View file

@ -632,7 +632,7 @@ paths:
operationId: "UpdateBucket"
summary: "Update a bucket"
description: |
All fields (`websiteAccess` and `quotas`) are optionnal.
All fields (`websiteAccess` and `quotas`) are optional.
If they are present, the corresponding modifications are applied to the bucket, otherwise nothing is changed.
In `websiteAccess`: if `enabled` is `true`, `indexDocument` must be specified.

View file

@ -0,0 +1,24 @@
<!DOCTYPE html>
<html>
<head>
<title>Garage Adminstration API v0</title>
<!-- needed for adaptive design -->
<meta charset="utf-8"/>
<meta name="viewport" content="width=device-width, initial-scale=1">
<link href="./css/redoc.css" rel="stylesheet">
<!--
Redoc doesn't change outer page styles
-->
<style>
body {
margin: 0;
padding: 0;
}
</style>
</head>
<body>
<redoc spec-url='./garage-admin-v1.yml'></redoc>
<script src="./redoc.standalone.js"> </script>
</body>
</html>

1362
doc/api/garage-admin-v1.yml Normal file

File diff suppressed because it is too large Load diff

View file

@ -1,6 +1,6 @@
+++
title = "Build your own app"
weight = 4
weight = 40
sort_by = "weight"
template = "documentation.html"
+++

View file

@ -37,30 +37,84 @@ import (
"context"
"fmt"
"os"
"strings"
garage "git.deuxfleurs.fr/garage-sdk/garage-admin-sdk-golang"
)
func main() {
// Set Host and other parameters
// Initialization
configuration := garage.NewConfiguration()
configuration.Host = "127.0.0.1:3903"
// We can now generate a client
client := garage.NewAPIClient(configuration)
// Authentication is handled through the context pattern
ctx := context.WithValue(context.Background(), garage.ContextAccessToken, "s3cr3t")
// Send a request
resp, r, err := client.NodesApi.GetNodes(ctx).Execute()
if err != nil {
fmt.Fprintf(os.Stderr, "Error when calling `NodesApi.GetNodes``: %v\n", err)
fmt.Fprintf(os.Stderr, "Full HTTP response: %v\n", r)
// Nodes
fmt.Println("--- nodes ---")
nodes, _, _ := client.NodesApi.GetNodes(ctx).Execute()
fmt.Fprintf(os.Stdout, "First hostname: %v\n", nodes.KnownNodes[0].Hostname)
capa := int64(1000000000)
change := []garage.NodeRoleChange{
garage.NodeRoleChange{NodeRoleUpdate: &garage.NodeRoleUpdate {
Id: *nodes.KnownNodes[0].Id,
Zone: "dc1",
Capacity: *garage.NewNullableInt64(&capa),
Tags: []string{ "fast", "amd64" },
}},
}
staged, _, _ := client.LayoutApi.AddLayout(ctx).NodeRoleChange(change).Execute()
msg, _, _ := client.LayoutApi.ApplyLayout(ctx).LayoutVersion(*garage.NewLayoutVersion(staged.Version + 1)).Execute()
fmt.Printf(strings.Join(msg.Message, "\n")) // Layout configured
// Process the response
fmt.Fprintf(os.Stdout, "Target hostname: %v\n", resp.KnownNodes[resp.Node].Hostname)
health, _, _ := client.NodesApi.GetHealth(ctx).Execute()
fmt.Printf("Status: %s, nodes: %v/%v, storage: %v/%v, partitions: %v/%v\n", health.Status, health.ConnectedNodes, health.KnownNodes, health.StorageNodesOk, health.StorageNodes, health.PartitionsAllOk, health.Partitions)
// Key
fmt.Println("\n--- key ---")
key := "openapi-key"
keyInfo, _, _ := client.KeyApi.AddKey(ctx).AddKeyRequest(garage.AddKeyRequest{Name: *garage.NewNullableString(&key) }).Execute()
defer client.KeyApi.DeleteKey(ctx).Id(*keyInfo.AccessKeyId).Execute()
fmt.Printf("AWS_ACCESS_KEY_ID=%s\nAWS_SECRET_ACCESS_KEY=%s\n", *keyInfo.AccessKeyId, *keyInfo.SecretAccessKey.Get())
id := *keyInfo.AccessKeyId
canCreateBucket := true
updateKeyRequest := *garage.NewUpdateKeyRequest()
updateKeyRequest.SetName("openapi-key-updated")
updateKeyRequest.SetAllow(garage.UpdateKeyRequestAllow { CreateBucket: &canCreateBucket })
update, _, _ := client.KeyApi.UpdateKey(ctx).Id(id).UpdateKeyRequest(updateKeyRequest).Execute()
fmt.Printf("Updated %v with key name %v\n", *update.AccessKeyId, *update.Name)
keyList, _, _ := client.KeyApi.ListKeys(ctx).Execute()
fmt.Printf("Keys count: %v\n", len(keyList))
// Bucket
fmt.Println("\n--- bucket ---")
global_name := "global-ns-openapi-bucket"
local_name := "local-ns-openapi-bucket"
bucketInfo, _, _ := client.BucketApi.CreateBucket(ctx).CreateBucketRequest(garage.CreateBucketRequest{
GlobalAlias: &global_name,
LocalAlias: &garage.CreateBucketRequestLocalAlias {
AccessKeyId: keyInfo.AccessKeyId,
Alias: &local_name,
},
}).Execute()
defer client.BucketApi.DeleteBucket(ctx).Id(*bucketInfo.Id).Execute()
fmt.Printf("Bucket id: %s\n", *bucketInfo.Id)
updateBucketRequest := *garage.NewUpdateBucketRequest()
website := garage.NewUpdateBucketRequestWebsiteAccess()
website.SetEnabled(true)
website.SetIndexDocument("index.html")
website.SetErrorDocument("errors/4xx.html")
updateBucketRequest.SetWebsiteAccess(*website)
quotas := garage.NewUpdateBucketRequestQuotas()
quotas.SetMaxSize(1000000000)
quotas.SetMaxObjects(999999999)
updateBucketRequest.SetQuotas(*quotas)
updatedBucket, _, _ := client.BucketApi.UpdateBucket(ctx).Id(*bucketInfo.Id).UpdateBucketRequest(updateBucketRequest).Execute()
fmt.Printf("Bucket %v website activation: %v\n", *updatedBucket.Id, *updatedBucket.WebsiteAccess)
bucketList, _, _ := client.BucketApi.ListBuckets(ctx).Execute()
fmt.Printf("Bucket count: %v\n", len(bucketList))
}
```

View file

@ -31,9 +31,9 @@ npm install --save git+https://git.deuxfleurs.fr/garage-sdk/garage-admin-sdk-js.
A short example:
```javascript
const garage = require('garage_administration_api_v0garage_v0_8_0');
const garage = require('garage_administration_api_v1garage_v0_9_0');
const api = new garage.ApiClient("http://127.0.0.1:3903/v0");
const api = new garage.ApiClient("http://127.0.0.1:3903/v1");
api.authentications['bearerAuth'].accessToken = "s3cr3t";
const [node, layout, key, bucket] = [

View file

@ -23,7 +23,7 @@ client = minio.Minio(
"GKyourapikey",
"abcd[...]1234",
# Force the region, this is specific to garage
region="region",
region="garage",
)
```
@ -80,7 +80,7 @@ from garage_admin_sdk.apis import *
from garage_admin_sdk.models import *
configuration = garage_admin_sdk.Configuration(
host = "http://localhost:3903/v0",
host = "http://localhost:3903/v1",
access_token = "s3cr3t"
)
@ -94,13 +94,14 @@ print(f"running garage {status.garage_version}, node_id {status.node}")
# Change layout of this node
current = layout.get_layout()
layout.add_layout({
status.node: NodeClusterInfo(
layout.add_layout([
NodeRoleChange(
id = status.node,
zone = "dc1",
capacity = 1,
capacity = 1000000000,
tags = [ "dev" ],
)
})
])
layout.apply_layout(LayoutVersion(
version = current.version + 1
))

View file

@ -1,6 +1,6 @@
+++
title = "Existing integrations"
weight = 3
weight = 30
sort_by = "weight"
template = "documentation.html"
+++

View file

@ -11,6 +11,7 @@ In this section, we cover the following web applications:
| [Peertube](#peertube) | ✅ | Supported with the website endpoint, proxifying private videos unsupported |
| [Mastodon](#mastodon) | ✅ | Natively supported |
| [Matrix](#matrix) | ✅ | Tested with `synapse-s3-storage-provider` |
| [ejabberd](#ejabberd) | ✅ | `mod_s3_upload` |
| [Pixelfed](#pixelfed) | ❓ | Not yet tested |
| [Pleroma](#pleroma) | ❓ | Not yet tested |
| [Lemmy](#lemmy) | ✅ | Supported with pict-rs |
@ -36,7 +37,7 @@ Second, we suppose you have created a key and a bucket.
As a reminder, you can create a key for your nextcloud instance as follow:
```bash
garage key new --name nextcloud-key
garage key create nextcloud-key
```
Keep the Key ID and the Secret key in a pad, they will be needed later.
@ -79,6 +80,53 @@ To test your new configuration, just reload your Nextcloud webpage and start sen
*External link:* [Nextcloud Documentation > Primary Storage](https://docs.nextcloud.com/server/latest/admin_manual/configuration_files/primary_storage.html)
#### SSE-C encryption (since Garage v1.0)
Since version 1.0, Garage supports server-side encryption with customer keys
(SSE-C). In this mode, Garage is responsible for encrypting and decrypting
objects, but it does not store the encryption key itself. The encryption key
should be provided by Nextcloud upon each request. This mode of operation is
supported by Nextcloud and it has successfully been tested together with
Garage.
To enable SSE-C encryption:
1. Make sure your Garage server is accessible via SSL through a reverse proxy
such as Nginx, and that it is using a valid public certificate (Nextcloud
might be able to connect to an S3 server that is using a self-signed
certificate, but you will lose many hours while trying, so don't).
Configure values for `use_ssl` and `port` accordingly in your `config.php`
file.
2. Generate an encryption key using the following command:
```
openssl rand -base64 32
```
Make sure to keep this key **secret**!
3. Add the encryption key in your `config.php` file as follows:
```php
<?php
$CONFIG = array(
'objectstore' => [
'class' => '\\OC\\Files\\ObjectStore\\S3',
'arguments' => [
...
'sse_c_key' => 'exampleencryptionkeyLbU+5fKYQcVoqnn+RaIOXgo=',
...
],
],
```
Nextcloud will now make Garage encrypt files at rest in the storage bucket.
These files will not be readable by an S3 client that has credentials to the
bucket but doesn't also know the secret encryption key.
### External Storage
**From the GUI.** Activate the "External storage support" app from the "Applications" page (click on your account icon on the top right corner of your screen to display the menu). Go to your parameters page (also located below your account icon). Click on external storage (or the corresponding translation in your language).
@ -138,14 +186,14 @@ a reasonable trade-off for some instances.
Create a key for Peertube:
```bash
garage key new --name peertube-key
garage key create peertube-key
```
Keep the Key ID and the Secret key in a pad, they will be needed later.
We need two buckets, one for normal videos (named peertube-video) and one for webtorrent videos (named peertube-playlist).
```bash
garage bucket create peertube-video
garage bucket create peertube-videos
garage bucket create peertube-playlist
```
@ -215,7 +263,7 @@ object_storage:
# Same settings but for webtorrent videos
videos:
bucket_name: 'peertube-video'
bucket_name: 'peertube-videos'
prefix: ''
# You must fill this field to make Peertube use our reverse proxy/website logic
base_url: 'http://peertube-videos.web.garage.localhost'
@ -244,7 +292,7 @@ with average object size ranging from 50 KB to 150 KB.
As such, your Garage cluster should be configured appropriately for good performance:
- use Garage v0.8.0 or higher with the [LMDB database engine](@documentation/reference-manual/configuration.md#db-engine-since-v0-8-0).
With the default Sled database engine, your database could quickly end up taking tens of GB of disk space.
Older versions of Garage used the Sled database engine which had issues, such as databases quickly ending up taking tens of GB of disk space.
- the Garage database should be stored on a SSD
### Creating your bucket
@ -252,7 +300,7 @@ As such, your Garage cluster should be configured appropriately for good perform
This is the usual Garage setup:
```bash
garage key new --name mastodon-key
garage key create mastodon-key
garage bucket create mastodon-data
garage bucket allow mastodon-data --read --write --key mastodon-key
```
@ -287,6 +335,7 @@ From the [official Mastodon documentation](https://docs.joinmastodon.org/admin/t
```bash
$ RAILS_ENV=production bin/tootctl media remove --days 3
$ RAILS_ENV=production bin/tootctl media remove --days 15 --prune-profiles
$ RAILS_ENV=production bin/tootctl media remove-orphans
$ RAILS_ENV=production bin/tootctl preview_cards remove --days 15
```
@ -305,8 +354,6 @@ Imports: 1.7 KB
Settings: 0 Bytes
```
Unfortunately, [old avatars and headers cannot currently be cleaned up](https://github.com/mastodon/mastodon/issues/9567).
### Migrating your data
Data migration should be done with an efficient S3 client.
@ -378,7 +425,7 @@ Supposing you have a working synapse installation, you can add the module with p
Now create a bucket and a key for your matrix instance (note your Key ID and Secret Key somewhere, they will be needed later):
```bash
garage key new --name matrix-key
garage key create matrix-key
garage bucket create matrix
garage bucket allow matrix --read --write --key matrix-key
```
@ -420,7 +467,7 @@ Now we can write a simple script (eg `~/.local/bin/matrix-cache-gc`):
## CONFIGURATION ##
AWS_ACCESS_KEY_ID=GKxxx
AWS_SECRET_ACCESS_KEY=xxxx
S3_ENDPOINT=http://localhost:3900
AWS_ENDPOINT_URL=http://localhost:3900
S3_BUCKET=matrix
MEDIA_STORE=/var/lib/matrix-synapse/media
PG_USER=matrix
@ -441,7 +488,7 @@ EOF
s3_media_upload update-db 1d
s3_media_upload --no-progress check-deleted $MEDIA_STORE
s3_media_upload --no-progress upload $MEDIA_STORE $S3_BUCKET --delete --endpoint-url $S3_ENDPOINT
s3_media_upload --no-progress upload $MEDIA_STORE $S3_BUCKET --delete --endpoint-url $AWS_ENDPOINT_URL
```
This script will list all the medias that were not accessed in the 24 hours according to your database.
@ -474,6 +521,52 @@ And add a new line. For example, to run it every 10 minutes:
*External link:* [matrix-media-repo Documentation > S3](https://docs.t2bot.io/matrix-media-repo/configuration/s3-datastore.html)
## ejabberd
ejabberd is an XMPP server implementation which, with the `mod_s3_upload`
module in the [ejabberd-contrib](https://github.com/processone/ejabberd-contrib)
repository, can be integrated to store chat media files in Garage.
For uploads, this module leverages presigned URLs - this allows XMPP clients to
directly send media to Garage. Receiving clients then retrieve this media
through the [static website](@/documentation/cookbook/exposing-websites.md)
functionality.
As the data itself is publicly accessible to someone with knowledge of the
object URL - users are recommended to use
[E2EE](@/documentation/cookbook/encryption.md) to protect this data-at-rest
from unauthorized access.
Install the module with:
```bash
ejabberdctl module_install mod_s3_upload
```
Create the required key and bucket with:
```bash
garage key new --name ejabberd
garage bucket create objects.xmpp-server.fr
garage bucket allow objects.xmpp-server.fr --read --write --key ejabberd
garage bucket website --allow objects.xmpp-server.fr
```
The module can then be configured with:
```
mod_s3_upload:
#bucket_url: https://objects.xmpp-server.fr.my-garage-instance.mydomain.tld
bucket_url: https://my-garage-instance.mydomain.tld/objects.xmpp-server.fr
access_key_id: GK...
access_key_secret: ...
region: garage
download_url: https://objects.xmpp-server.fr
```
Other configuration options can be found in the
[configuration YAML file](https://github.com/processone/ejabberd-contrib/blob/master/mod_s3_upload/conf/mod_s3_upload.yml).
## Pixelfed
[Pixelfed Technical Documentation > Configuration](https://docs.pixelfed.org/technical-documentation/env.html#filesystem)
@ -539,7 +632,7 @@ secret_key = 'abcdef0123456789...'
```
PICTRS__STORE__TYPE=object_storage
PICTRS__STORE__ENDPOINT=http:/my-garage-instance.mydomain.tld:3900
PICTRS__STORE__ENDPOINT=http://my-garage-instance.mydomain.tld:3900
PICTRS__STORE__BUCKET_NAME=pictrs-data
PICTRS__STORE__REGION=garage
PICTRS__STORE__ACCESS_KEY=GK...

View file

@ -54,9 +54,9 @@ how to configure this.
Create your key and bucket:
```bash
garage key new my-key
garage bucket create backup
garage bucket allow backup --read --write --key my-key
garage key create my-key
garage bucket create backups
garage bucket allow backups --read --write --key my-key
```
Then register your Key ID and Secret key in your environment:
@ -105,6 +105,7 @@ restic restore 79766175 --target /var/lib/postgresql
Restic has way more features than the ones presented here.
You can discover all of them by accessing its documentation from the link below.
Files on Android devices can also be backed up with [restic-android](https://github.com/lhns/restic-android).
*External links:* [Restic Documentation > Amazon S3](https://restic.readthedocs.io/en/stable/030_preparing_a_new_repo.html#amazon-s3)

View file

@ -70,16 +70,17 @@ Then a file named `~/.aws/config` and put:
```toml
[default]
region=garage
endpoint_url=http://127.0.0.1:3900
```
Now, supposing Garage is listening on `http://127.0.0.1:3900`, you can list your buckets with:
```bash
aws --endpoint-url http://127.0.0.1:3900 s3 ls
aws s3 ls
```
Passing the `--endpoint-url` parameter to each command is annoying but AWS developers do not provide a corresponding configuration entry.
As a workaround, you can redefine the aws command by editing the file `~/.bashrc`:
If you're using awscli `<1.29.0` or `<2.13.0`, you need to pass `--endpoint-url` to each CLI invocation explicitly.
As a workaround, you can redefine the aws command by editing the file `~/.bashrc` in this case:
```
function aws { command aws --endpoint-url http://127.0.0.1:3900 $@ ; }
@ -258,7 +259,7 @@ duck --delete garage:/my-files/an-object.txt
## WinSCP (libs3) {#winscp}
*You can find instructions on how to use the GUI in french [in our wiki](https://wiki.deuxfleurs.fr/fr/Guide/Garage/WinSCP).*
*You can find instructions on how to use the GUI in french [in our wiki](https://guide.deuxfleurs.fr/prise_en_main/winscp/).*
How to use `winscp.com`, the CLI interface of WinSCP:

View file

@ -23,7 +23,7 @@ You can configure a different target for each data type (check `[lfs]` and `[att
Let's start by creating a key and a bucket (your key id and secret will be needed later, keep them somewhere):
```bash
garage key new --name gitea-key
garage key create gitea-key
garage bucket create gitea
garage bucket allow gitea --read --write --key gitea-key
```
@ -118,7 +118,7 @@ through another support, like a git repository.
As a first step, we will need to create a bucket on Garage and enabling website access on it:
```bash
garage key new --name nix-key
garage key create nix-key
garage bucket create nix.example.com
garage bucket allow nix.example.com --read --write --key nix-key
garage bucket website nix.example.com --allow

View file

@ -1,7 +1,7 @@
+++
title="Cookbook"
template = "documentation.html"
weight = 2
weight = 20
sort_by = "weight"
+++
@ -37,7 +37,3 @@ This chapter could also be referred as "Tutorials" or "Best practices".
- **[Monitoring Garage](@/documentation/cookbook/monitoring.md)** This page
explains the Prometheus metrics available for monitoring the Garage
cluster/nodes.
- **[Recovering from failures](@/documentation/cookbook/recovering.md):** Garage's first selling point is resilience
to hardware failures. This section explains how to recover from such a failure in the
best possible way.

View file

@ -7,10 +7,23 @@ Garage is also available in binary packages on:
## Alpine Linux
If you use Alpine Linux, you can simply install the
[garage](https://pkgs.alpinelinux.org/packages?name=garage) package from the
Alpine Linux repositories (available since v3.17):
```bash
apk install garage
apk add garage
```
The default configuration file is installed to `/etc/garage.toml`. You can run
Garage using: `rc-service garage start`. If you don't specify `rpc_secret`, it
will be automatically replaced with a random string on the first start.
Please note that this package is built without Consul discovery, Kubernetes
discovery, OpenTelemetry exporter, and K2V features (K2V will be enabled once
it's stable).
## Arch Linux
Garage is available in the [AUR](https://aur.archlinux.org/packages/garage).

View file

@ -0,0 +1,139 @@
+++
title = "Encryption"
weight = 50
+++
Encryption is a recurring subject when discussing Garage.
Garage does not handle data encryption by itself, but many things can
already be done with Garage's current feature set and the existing ecosystem.
This page takes a high level approach to security in general and data encryption
in particular.
# Examining your need for encryption
- Why do you want encryption in Garage?
- What is your threat model? What are you fearing?
- A stolen HDD?
- A curious administrator?
- A malicious administrator?
- A remote attacker?
- etc.
- What services do you want to protect with encryption?
- An existing application? Which one? (eg. Nextcloud)
- An application that you are writing
- Any expertise you may have on the subject
This page explains what Garage provides, and how you can improve the situation by yourself
by adding encryption at different levels.
We would be very curious to know your needs and thougs about ideas such as
encryption practices and things like key management, as we want Garage to be a
serious base platform for the developpment of secure, encrypted applications.
Do not hesitate to come talk to us if you have any thoughts or questions on the
subject.
# Capabilities provided by Garage
## Traffic is encrypted between Garage nodes
RPCs between Garage nodes are encrypted. More specifically, contrary to many
distributed software, it is impossible in Garage to have clear-text RPC. We
use the [kuska handshake](https://github.com/Kuska-ssb/handshake) library which
implements a protocol that has been clearly reviewed, Secure ScuttleButt's
Secret Handshake protocol. This is why setting a `rpc_secret` is mandatory,
and that's also why your nodes have super long identifiers.
## HTTP API endpoints provided by Garage are in clear text
Adding TLS support built into Garage is not currently planned.
## Garage stores data in plain text on the filesystem or encrypted using customer keys (SSE-C)
For standard S3 API requests, Garage does not encrypt data at rest by itself.
For the most generic at rest encryption of data, we recommend setting up your
storage partitions on encrypted LUKS devices.
If you are developping your own client software that makes use of S3 storage,
we recommend implementing data encryption directly on the client side and never
transmitting plaintext data to Garage. This makes it easy to use an external
untrusted storage provider if necessary.
Garage does support [SSE-C
encryption](https://docs.aws.amazon.com/AmazonS3/latest/userguide/ServerSideEncryptionCustomerKeys.html),
an encryption mode of Amazon S3 where data is encrypted at rest using
encryption keys given by the client. The encryption keys are passed to the
server in a header in each request, to encrypt or decrypt data at the moment of
reading or writing. The server discards the key as soon as it has finished
using it for the request. This mode allows the data to be encrypted at rest by
Garage itself, but it requires support in the client software. It is also not
adapted to a model where the server is not trusted or assumed to be
compromised, as the server can easily know the encryption keys. Note however
that when using SSE-C encryption, the only Garage node that knows the
encryption key passed in a given request is the node to which the request is
directed (which can be a gateway node), so it is easy to have untrusted nodes
in the cluster as long as S3 API requests containing SSE-C encryption keys are
not directed to them.
Implementing automatic data encryption directly in Garage without client-side
management of keys (something like
[SSE-S3](https://docs.aws.amazon.com/AmazonS3/latest/userguide/UsingServerSideEncryption.html))
could make things simpler for end users that don't want to setup LUKS, but also
raises many more questions, especially around key management: for encryption of
data, where could Garage get the encryption keys from? If we encrypt data but
keep the keys in a plaintext file next to them, it's useless. We probably don't
want to have to manage secrets in Garage as it would be very hard to do in a
secure way. At the time of speaking, there are no plans to implement this in
Garage.
# Adding data encryption using external tools
## Encrypting traffic between a Garage node and your client
You have multiple options to have encryption between your client and a node:
- Setup a reverse proxy with TLS / ACME / Let's encrypt
- Setup a Garage gateway locally, and only contact the garage daemon on `localhost`
- Only contact your Garage daemon over a secure, encrypted overlay network such as Wireguard
## Encrypting data at rest
Protects against the following threats:
- Stolen HDD
Crucially, does not protect againt malicious sysadmins or remote attackers that
might gain access to your servers.
Methods include full-disk encryption with tools such as LUKS.
## Encrypting data on the client side
Protects againt the following threats:
- A honest-but-curious administrator
- A malicious administrator that tries to corrupt your data
- A remote attacker that can read your server's data
Implementations are very specific to the various applications. Examples:
- Matrix: uses the OLM protocol for E2EE of user messages. Media files stored
in Matrix are probably encrypted using symmetric encryption, with a key that is
distributed in the end-to-end encrypted message that contains the link to the object.
- XMPP: clients normally support either OMEMO / OpenPGP for the E2EE of user
messages. Media files are encrypted per
[XEP-0454](https://xmpp.org/extensions/xep-0454.html).
- Aerogramme: use the user's password as a key to decrypt data in the user's bucket
- Cyberduck: comes with support for
[Cryptomator](https://docs.cyberduck.io/cryptomator/) which allows users to
create client-side vaults to encrypt files in before they are uploaded to a
cloud storage endpoint.

View file

@ -38,7 +38,7 @@ Our website serving logic is as follow:
Now we need to infer the URL of your website through your bucket name.
Let assume:
- we set `root_domain = ".web.example.com"` in `garage.toml` ([ref](@/documentation/reference-manual/configuration.md#root_domain))
- we set `root_domain = ".web.example.com"` in `garage.toml` ([ref](@/documentation/reference-manual/configuration.md#web_root_domain))
- our bucket name is `garagehq.deuxfleurs.fr`.
Our bucket will be served if the Host field matches one of these 2 values (the port is ignored):

View file

@ -90,6 +90,6 @@ The following feature flags are available in v0.8.0:
| `kubernetes-discovery` | optional | Enable automatic registration and discovery<br>of cluster nodes through the Kubernetes API |
| `metrics` | *by default* | Enable collection of metrics in Prometheus format on the admin API |
| `telemetry-otlp` | optional | Enable collection of execution traces using OpenTelemetry |
| `sled` | *by default* | Enable using Sled to store Garage's metadata |
| `lmdb` | optional | Enable using LMDB to store Garage's metadata |
| `sqlite` | optional | Enable using Sqlite3 to store Garage's metadata |
| `syslog` | optional | Enable logging to Syslog |
| `lmdb` | *by default* | Enable using LMDB to store Garage's metadata |
| `sqlite` | *by default* | Enable using Sqlite3 to store Garage's metadata |

View file

@ -18,7 +18,7 @@ api_bind_addr = "0.0.0.0:3903"
```
This will allow anyone to scrape Prometheus metrics by fetching
`http://localhost:3093/metrics`. If you want to restrict access
`http://localhost:3903/metrics`. If you want to restrict access
to the exported metrics, set the `metrics_token` configuration value
to a bearer token to be used when fetching the metrics endpoint.
@ -49,9 +49,5 @@ add the following lines in your Prometheus scrape config:
To visualize the scraped data in Grafana,
you can either import our [Grafana dashboard for Garage](https://git.deuxfleurs.fr/Deuxfleurs/garage/raw/branch/main/script/telemetry/grafana-garage-dashboard-prometheus.json)
or make your own.
We detail below the list of exposed metrics and their meaning.
## List of exported metrics
See our [dedicated page](@/documentation/reference-manual/monitoring.md) in the Reference manual section.
The list of exported metrics is available on our [dedicated page](@/documentation/reference-manual/monitoring.md) in the Reference manual section.

View file

@ -19,9 +19,10 @@ To run a real-world deployment, make sure the following conditions are met:
- You have at least three machines with sufficient storage space available.
- Each machine has a public IP address which is reachable by other machines. It
is highly recommended that you use IPv6 for this end-to-end connectivity. If
IPv6 is not available, then using a mesh VPN such as
- Each machine has an IP address which makes it directly reachable by all other machines.
In many cases, nodes will be behind a NAT and will not each have a public
IPv4 addresses. In this case, is recommended that you use IPv6 for this
end-to-end connectivity if it is available. Otherwise, using a mesh VPN such as
[Nebula](https://github.com/slackhq/nebula) or
[Yggdrasil](https://yggdrasil-network.github.io/) are approaches to consider
in addition to building out your own VPN tunneling.
@ -42,7 +43,7 @@ For our example, we will suppose the following infrastructure with IPv6 connecti
| Brussels | Mars | fc00:F::1 | 1.5 TB |
Note that Garage will **always** store the three copies of your data on nodes at different
locations. This means that in the case of this small example, the available capacity
locations. This means that in the case of this small example, the usable capacity
of the cluster is in fact only 1.5 TB, because nodes in Brussels can't store more than that.
This also means that nodes in Paris and London will be under-utilized.
To make better use of the available hardware, you should ensure that the capacity
@ -52,9 +53,9 @@ to store 2 TB of data in total.
### Best practices
- If you have fast dedicated networking between all your nodes, and are planing to store
very large files, bump the `block_size` configuration parameter to 10 MB
(`block_size = 10485760`).
- If you have reasonably fast networking between all your nodes, and are planing to store
mostly large files, bump the `block_size` configuration parameter to 10 MB
(`block_size = "10M"`).
- Garage stores its files in two locations: it uses a metadata directory to store frequently-accessed
small metadata items, and a data directory to store data blocks of uploaded objects.
@ -67,36 +68,42 @@ to store 2 TB of data in total.
EXT4 is not recommended as it has more strict limitations on the number of inodes,
which might cause issues with Garage when large numbers of objects are stored.
- If you only have an HDD and no SSD, it's fine to put your metadata alongside the data
on the same drive. Having lots of RAM for your kernel to cache the metadata will
help a lot with performance. Make sure to use the LMDB database engine,
instead of Sled, which suffers from quite bad performance degradation on HDDs.
Sled is still the default for legacy reasons, but is not recommended anymore.
- Servers with multiple HDDs are supported natively by Garage without resorting
to RAID, see [our dedicated documentation page](@/documentation/operations/multi-hdd.md).
- For the metadata storage, Garage does not do checksumming and integrity
verification on its own. If you are afraid of bitrot/data corruption,
put your metadata directory on a BTRFS partition. Otherwise, just use regular
EXT4 or XFS.
verification on its own, so it is better to use a robust filesystem such as
BTRFS or ZFS. Users have reported that when using the LMDB database engine
(the default), database files have a tendency of becoming corrupted after an
unclean shutdown (e.g. a power outage), so you should take regular snapshots
to be able to recover from such a situation. This can be done using Garage's
built-in automatic snapshotting (since v0.9.4), or by using filesystem level
snapshots. If you cannot do so, you might want to switch to Sqlite which is
more robust.
- Having a single server with several storage drives is currently not very well
supported in Garage ([#218](https://git.deuxfleurs.fr/Deuxfleurs/garage/issues/218)).
For an easy setup, just put all your drives in a RAID0 or a ZFS RAIDZ array.
If you're adventurous, you can try to format each of your disk as
a separate XFS partition, and then run one `garage` daemon per disk drive,
or use something like [`mergerfs`](https://github.com/trapexit/mergerfs) to merge
all your disks in a single union filesystem that spreads load over them.
- LMDB is the fastest and most tested database engine, but it has the following
weaknesses: 1/ data files are not architecture-independent, you cannot simply
move a Garage metadata directory between nodes running different architectures,
and 2/ LMDB is not suited for 32-bit platforms. Sqlite is a viable alternative
if any of these are of concern.
- If you only have an HDD and no SSD, it's fine to put your metadata alongside
the data on the same drive, but then consider your filesystem choice wisely
(see above). Having lots of RAM for your kernel to cache the metadata will
help a lot with performance. The default LMDB database engine is the most
tested and has good performance.
## Get a Docker image
Our docker image is currently named `dxflrs/garage` and is stored on the [Docker Hub](https://hub.docker.com/r/dxflrs/garage/tags?page=1&ordering=last_updated).
We encourage you to use a fixed tag (eg. `v0.8.0`) and not the `latest` tag.
For this example, we will use the latest published version at the time of the writing which is `v0.8.0` but it's up to you
We encourage you to use a fixed tag (eg. `v1.0.1`) and not the `latest` tag.
For this example, we will use the latest published version at the time of the writing which is `v1.0.1` but it's up to you
to check [the most recent versions on the Docker Hub](https://hub.docker.com/r/dxflrs/garage/tags?page=1&ordering=last_updated).
For example:
```
sudo docker pull dxflrs/garage:v0.8.0
sudo docker pull dxflrs/garage:v1.0.1
```
## Deploying and configuring Garage
@ -119,8 +126,9 @@ A valid `/etc/garage.toml` for our cluster would look as follows:
metadata_dir = "/var/lib/garage/meta"
data_dir = "/var/lib/garage/data"
db_engine = "lmdb"
metadata_auto_snapshot_interval = "6h"
replication_mode = "3"
replication_factor = 3
compression_level = 2
@ -144,6 +152,8 @@ Check the following for your configuration files:
- Make sure `rpc_public_addr` contains the public IP address of the node you are configuring.
This parameter is optional but recommended: if your nodes have trouble communicating with
one another, consider adding it.
Alternatively, you can also set `rpc_public_addr_subnet`, which can filter
the addresses announced to other peers to a specific subnet.
- Make sure `rpc_secret` is the same value on all nodes. It should be a 32-bytes hex-encoded secret key.
You can generate such a key with `openssl rand -hex 32`.
@ -161,12 +171,13 @@ docker run \
-v /etc/garage.toml:/etc/garage.toml \
-v /var/lib/garage/meta:/var/lib/garage/meta \
-v /var/lib/garage/data:/var/lib/garage/data \
dxflrs/garage:v0.8.0
dxflrs/garage:v1.0.1
```
It should be restarted automatically at each reboot.
Please note that we use host networking as otherwise Docker containers
can not communicate with IPv6.
With this command line, Garage should be started automatically at each boot.
Please note that we use host networking as otherwise the network indirection
added by Docker would prevent Garage nodes from communicating with one another
(especially if using IPv6).
If you want to use `docker-compose`, you may use the following `docker-compose.yml` file as a reference:
@ -174,7 +185,7 @@ If you want to use `docker-compose`, you may use the following `docker-compose.y
version: "3"
services:
garage:
image: dxflrs/garage:v0.8.0
image: dxflrs/garage:v1.0.1
network_mode: "host"
restart: unless-stopped
volumes:
@ -183,12 +194,14 @@ services:
- /var/lib/garage/data:/var/lib/garage/data
```
Upgrading between Garage versions should be supported transparently,
but please check the relase notes before doing so!
To upgrade, simply stop and remove this container and
start again the command with a new version of Garage.
If you wish to upgrade your cluster, make sure to read the corresponding
[documentation page](@/documentation/operations/upgrading.md) first, as well as
the documentation relevant to your version of Garage in the case of major
upgrades. With the containerized setup proposed here, the upgrade process
will require stopping and removing the existing container, and re-creating it
with the upgraded version.
## Controling the daemon
## Controlling the daemon
The `garage` binary has two purposes:
- it acts as a daemon when launched with `garage server`
@ -197,6 +210,12 @@ The `garage` binary has two purposes:
Ensure an appropriate `garage` binary (the same version as your Docker image) is available in your path.
If your configuration file is at `/etc/garage.toml`, the `garage` binary should work with no further change.
You can also use an alias as follows to use the Garage binary inside your docker container:
```bash
alias garage="docker exec -ti <container name> /garage"
```
You can test your `garage` CLI utility by running a simple command such as:
```bash
@ -240,7 +259,7 @@ You can then instruct nodes to connect to one another as follows:
Venus$ garage node connect 563e1ac825ee3323aa441e72c26d1030d6d4414aeb3dd25287c531e7fc2bc95d@[fc00:1::1]:3901
```
You don't nead to instruct all node to connect to all other nodes:
You don't need to instruct all node to connect to all other nodes:
nodes will discover one another transitively.
Now if your run `garage status` on any node, you should have an output that looks as follows:
@ -264,12 +283,12 @@ of a role that is assigned to each active cluster node.
For our example, we will suppose we have the following infrastructure
(Capacity, Identifier and Zone are specific values to Garage described in the following):
| Location | Name | Disk Space | `Capacity` | `Identifier` | `Zone` |
|----------|---------|------------|------------|--------------|--------------|
| Paris | Mercury | 1 TB | `10` | `563e` | `par1` |
| Paris | Venus | 2 TB | `20` | `86f0` | `par1` |
| London | Earth | 2 TB | `20` | `6814` | `lon1` |
| Brussels | Mars | 1.5 TB | `15` | `212f` | `bru1` |
| Location | Name | Disk Space | Identifier | Zone (`-z`) | Capacity (`-c`) |
|----------|---------|------------|------------|-------------|-----------------|
| Paris | Mercury | 1 TB | `563e` | `par1` | `1T` |
| Paris | Venus | 2 TB | `86f0` | `par1` | `2T` |
| London | Earth | 2 TB | `6814` | `lon1` | `2T` |
| Brussels | Mars | 1.5 TB | `212f` | `bru1` | `1.5T` |
#### Node identifiers
@ -291,6 +310,8 @@ garage status
It will display the IP address associated with each node;
from the IP address you will be able to recognize the node.
We will now use the `garage layout assign` command to configure the correct parameters for each node.
#### Zones
Zones are simply a user-chosen identifier that identify a group of server that are grouped together logically.
@ -300,29 +321,29 @@ In most cases, a zone will correspond to a geographical location (i.e. a datacen
Behind the scene, Garage will use zone definition to try to store the same data on different zones,
in order to provide high availability despite failure of a zone.
Zones are passed to Garage using the `-z` flag of `garage layout assign` (see below).
#### Capacity
Garage reasons on an abstract metric about disk storage that is named the *capacity* of a node.
The capacity configured in Garage must be proportional to the disk space dedicated to the node.
Garage needs to know the storage capacity (disk space) it can/should use on
each node, to be able to correctly balance data.
Capacity values must be **integers** but can be given any signification.
Here we chose that 1 unit of capacity = 100 GB.
Capacity values are expressed in bytes and are passed to Garage using the `-c` flag of `garage layout assign` (see below).
Note that the amount of data stored by Garage on each server may not be strictly proportional to
its capacity value, as Garage will priorize having 3 copies of data in different zones,
even if this means that capacities will not be strictly respected. For example in our above examples,
nodes Earth and Mars will always store a copy of everything each, and the third copy will
have 66% chance of being stored by Venus and 33% chance of being stored by Mercury.
#### Tags
You can add additional tags to nodes using the `-t` flag of `garage layout assign` (see below).
Tags have no specific meaning for Garage and can be used at your convenience.
#### Injecting the topology
Given the information above, we will configure our cluster as follow:
```bash
garage layout assign 563e -z par1 -c 10 -t mercury
garage layout assign 86f0 -z par1 -c 20 -t venus
garage layout assign 6814 -z lon1 -c 20 -t earth
garage layout assign 212f -z bru1 -c 15 -t mars
garage layout assign 563e -z par1 -c 1T -t mercury
garage layout assign 86f0 -z par1 -c 2T -t venus
garage layout assign 6814 -z lon1 -c 2T -t earth
garage layout assign 212f -z bru1 -c 1.5T -t mars
```
At this point, the changes in the cluster layout have not yet been applied.
@ -332,6 +353,7 @@ To show the new layout that will be applied, call:
garage layout show
```
Make sure to read carefully the output of `garage layout show`.
Once you are satisfied with your new layout, apply it with:
```bash
@ -339,7 +361,7 @@ garage layout apply
```
**WARNING:** if you want to use the layout modification commands in a script,
make sure to read [this page](@/documentation/reference-manual/layout.md) first.
make sure to read [this page](@/documentation/operations/layout.md) first.
## Using your Garage cluster
@ -349,5 +371,5 @@ and is covered in the [quick start guide](@/documentation/quick-start/_index.md)
Remember also that the CLI is self-documented thanks to the `--help` flag and
the `help` subcommand (e.g. `garage help`, `garage key --help`).
Configuring S3-compatible applicatiosn to interact with Garage
Configuring S3-compatible applications to interact with Garage
is covered in the [Integrations](@/documentation/connect/_index.md) section.

View file

@ -378,6 +378,47 @@ admin.garage.tld {
But at the same time, the `reverse_proxy` is very flexible.
For a production deployment, you should [read its documentation](https://caddyserver.com/docs/caddyfile/directives/reverse_proxy) as it supports features like DNS discovery of upstreams, load balancing with checks, streaming parameters, etc.
### Caching
Caddy can compiled with a
[cache plugin](https://github.com/caddyserver/cache-handler) which can be used
to provide a hot-cache at the webserver-level for static websites hosted by
Garage.
This can be configured as follows:
```caddy
# Caddy global configuration section
{
# Bare minimum configuration to enable cache.
order cache before rewrite
cache
#cache
# allowed_http_verbs GET
# default_cache_control public
# ttl 8h
#}
}
# Site specific section
https:// {
cache
#cache {
# timeout {
# backend 30s
# }
#}
reverse_proxy ...
}
```
Caching is a complicated subject, and the reader is encouraged to study the
available options provided by the plugin.
### On-demand TLS
Caddy supports a technique called
@ -428,3 +469,35 @@ https:// {
reverse_proxy localhost:3902 192.168.1.2:3902 example.tld:3902
}
```
More information on how this endpoint is implemented in Garage is available
in the [Admin API Reference](@/documentation/reference-manual/admin-api.md) page.
### Fileserver browser
Caddy's built-in
[file_server](https://caddyserver.com/docs/caddyfile/directives/file_server)
browser functionality can be extended with the
[caddy-fs-s3](https://github.com/sagikazarmark/caddy-fs-s3) module.
This can be configured to use Garage as a backend with the following
configuration:
```caddy
browse.garage.tld {
file_server {
fs s3 {
bucket test-bucket
region garage
endpoint https://s3.garage.tld
use_path_style
}
browse
}
}
```
Caddy must also be configured with the required `AWS_ACCESS_KEY_ID` and
`AWS_SECRET_ACCESS_KEY` environment variables to access the bucket.

View file

@ -33,7 +33,20 @@ NoNewPrivileges=true
WantedBy=multi-user.target
```
*A note on hardening: garage will be run as a non privileged user, its user id is dynamically allocated by systemd. It cannot access (read or write) home folders (/home, /root and /run/user), the rest of the filesystem can only be read but not written, only the path seen as /var/lib/garage is writable as seen by the service (mapped to /var/lib/private/garage on your host). Additionnaly, the process can not gain new privileges over time.*
**A note on hardening:** Garage will be run as a non privileged user, its user
id is dynamically allocated by systemd (set with `DynamicUser=true`). It cannot
access (read or write) home folders (`/home`, `/root` and `/run/user`), the
rest of the filesystem can only be read but not written, only the path seen as
`/var/lib/garage` is writable as seen by the service. Additionnaly, the process
can not gain new privileges over time.
For this to work correctly, your `garage.toml` must be set with
`metadata_dir=/var/lib/garage/meta` and `data_dir=/var/lib/garage/data`. This
is mandatory to use the DynamicUser hardening feature of systemd, which
autocreates these directories as virtual mapping. If the directory
`/var/lib/garage` already exists before starting the server for the first time,
the systemd service might not start correctly. Note that in your host
filesystem, Garage data will be held in `/var/lib/private/garage`.
To start the service then automatically enable it at boot:

View file

@ -1,6 +1,6 @@
+++
title = "Design"
weight = 6
weight = 70
sort_by = "weight"
template = "documentation.html"
+++

View file

@ -42,15 +42,28 @@ locations. They use Garage themselves for the following tasks:
- As a [Matrix media backend](https://github.com/matrix-org/synapse-s3-storage-provider)
- To store personal data and shared documents through [Bagage](https://git.deuxfleurs.fr/Deuxfleurs/bagage), a homegrown WebDav-to-S3 proxy
- In the Drone continuous integration platform to store task logs
- As a Nix binary cache
- As a backup target using `rclone`
- To store personal data and shared documents through [Bagage](https://git.deuxfleurs.fr/Deuxfleurs/bagage), a homegrown WebDav-to-S3 and SFTP-to-S3 proxy
- As a backup target using `rclone` and `restic`
The Deuxfleurs Garage cluster is a multi-site cluster currently composed of
4 nodes in 2 physical locations. In the future it will be expanded to at
least 3 physical locations to fully exploit Garage's potential for high
availability.
9 nodes in 3 physical locations.
### Triplebit
[Triplebit](https://www.triplebit.org) is a non-profit hosting provider and
ISP focused on improving access to privacy-related services. They use
Garage themselves for the following tasks:
- Hosting of their homepage, [privacyguides.org](https://www.privacyguides.org/), and various other static sites
- As a Mastodon object storage backend for [mstdn.party](https://mstdn.party/) and [mstdn.plus](https://mstdn.plus/)
- As a PeerTube storage backend for [neat.tube](https://neat.tube/)
- As a [Matrix media backend](https://github.com/matrix-org/synapse-s3-storage-provider)
Triplebit's Garage cluster is a multi-site cluster currently composed of
10 nodes in 3 physical locations.

View file

@ -61,7 +61,7 @@ Garage prioritizes which nodes to query according to a few criteria:
For further reading on the cluster structure look at the [gateway](@/documentation/cookbook/gateways.md)
and [cluster layout management](@/documentation/reference-manual/layout.md) pages.
and [cluster layout management](@/documentation/operations/layout.md) pages.
## Garbage collection
@ -97,7 +97,7 @@ delete a tombstone, the following condition has to be met:
superseeded by the tombstone. This ensures that deleting the tombstone is
safe and that no deleted value will come back in the system.
Garage makes use of Sled's atomic operations (such as compare-and-swap and
Garage uses atomic database operations (such as compare-and-swap and
transactions) to ensure that only tombstones that have been correctly
propagated to other nodes are ever deleted from the local entry tree.

View file

@ -67,7 +67,7 @@ Pithos has been abandonned and should probably not used yet, in the following we
Pithos was relying as a S3 proxy in front of Cassandra (and was working with Scylla DB too).
From its designers' mouth, storing data in Cassandra has shown its limitations justifying the project abandonment.
They built a closed-source version 2 that does not store blobs in the database (only metadata) but did not communicate further on it.
We considered there v2's design but concluded that it does not fit both our *Self-contained & lightweight* and *Simple* properties. It makes the development, the deployment and the operations more complicated while reducing the flexibility.
We considered their v2's design but concluded that it does not fit both our *Self-contained & lightweight* and *Simple* properties. It makes the development, the deployment and the operations more complicated while reducing the flexibility.
**[Riak CS](https://docs.riak.com/riak/cs/2.1.1/index.html):**
*Not written yet*

View file

@ -1,6 +1,6 @@
+++
title = "Development"
weight = 7
weight = 80
sort_by = "weight"
template = "documentation.html"
+++

View file

@ -25,7 +25,7 @@ git clone https://git.deuxfleurs.fr/Deuxfleurs/garage
cd garage
```
*Optionnaly, you can use our nix.conf file to speed up compilations:*
*Optionally, you can use our nix.conf file to speed up compilations:*
```bash
sudo mkdir -p /etc/nix
@ -80,7 +80,7 @@ nix-build \
--git_version $(git rev-parse HEAD)
```
*The result is located in `result/bin`. You can pass arguments to cross compile: check `.drone.yml` for examples.*
*The result is located in `result/bin`. You can pass arguments to cross compile: check `.woodpecker/release.yml` for examples.*
If you modify a `Cargo.toml` or regenerate any `Cargo.lock`, you must run `cargo2nix`:

View file

@ -81,12 +81,9 @@ Our cache will be checked.
- http://www.lpenz.org/articles/nixchannel/index.html
## Drone
## Woodpecker
Do not try to set a build as trusted from the interface or the CLI tool,
your request would be ignored. Instead, directly edit the database (table `repos`, column `repo_trusted`).
Drone can do parallelism both at the step and the pipeline level. At the step level, parallelism is restricted to the same runner.
Woodpecker can do parallelism both at the step and the pipeline level. At the step level, parallelism is restricted to the same runner.
## Building Docker containers
@ -99,3 +96,4 @@ We were:
- Unable to use the kaniko container provided by Google as we can't run arbitrary logic: we need to put our secret in .docker/config.json.
Finally we chose to build kaniko through nix and use it in a `nix-shell`.
We then switched to using kaniko from nixpkgs when it was packaged.

View file

@ -42,7 +42,7 @@ and the docker containers on Docker Hub.
## Automation
We automated our release process with Nix and Drone to make it more reliable.
We automated our release process with Nix and Woodpecker to make it more reliable.
Here we describe how we have done in case you want to debug or improve it.
### Caching build steps
@ -62,52 +62,31 @@ Sending to the cache is done through `nix copy`, for example:
nix copy --to 's3://nix?endpoint=garage.deuxfleurs.fr&region=garage&secret-key=/etc/nix/signing-key.sec' result
```
*Note that you need the signing key. In our case, it is stored as a secret in Drone.*
*The signing key possessed by the Garage maintainers is required to update the Nix cache.*
The previous command will only send the built packet and not its dependencies.
To send its dependency, a tool named `nix-copy-closure` has been created but it is not compatible with the S3 protocol.
Instead, you can use the following commands to list all the runtime dependencies:
The previous command will only send the built package and not its dependencies.
In the case of our CI pipeline, we want to cache all intermediate build steps
as well. This can be done using this quite involved command (here as an example
for the `pkgs.amd64.relase` package):
```bash
nix copy \
--to 's3://nix?endpoint=garage.deuxfleurs.fr&region=garage&secret-key=/etc/nix/signing-key.sec' \
$(nix-store -qR result/)
nix copy -j8 \
--to 's3://nix?endpoint=garage.deuxfleurs.fr&region=garage&secret-key=/etc/nix/nix-signing-key.sec' \
$(nix path-info pkgs.amd64.release --file default.nix --derivation --recursive | sed 's/\.drv$/.drv^*/')
```
*We could also write this expression with xargs but this tool is not available in our container.*
This command will simultaneously build all of the required Nix paths (using at
most 8 parallel Nix builder jobs) and send the resulting objects to the cache.
But in certain cases, we want to cache compile time dependencies also.
For example, the Nix project does not provide binaries for cross compiling to i686 and thus we need to compile gcc on our own.
We do not want to compile gcc each time, so even if it is a compile time dependency, we want to cache it.
This time, the command is a bit more involved:
```bash
nix copy --to \
's3://nix?endpoint=garage.deuxfleurs.fr&region=garage&secret-key=/etc/nix/signing-key.sec' \
$(nix-store -qR --include-outputs \
$(nix-instantiate))
```
This is the command we use in our CI as we expect the final binary to change, so we mainly focus on
caching our development dependencies.
*Currently there is no automatic garbage collection of the cache: we should monitor its growth.
Hopefully, we can erase it totally without breaking any build, the next build will only be slower.*
In practise, we concluded that we do not want to cache all the compilation dependencies.
Instead, we want to cache the toolchain we use to build Garage each time we change it.
So we removed from Drone any automatic update of the cache and instead handle them manually with:
This can be run for all the Garage packages we build using the following command:
```
source ~/.awsrc
nix-shell --run 'refresh_toolchain'
nix-shell --attr cache --run 'refresh_cache'
```
Internally, it will run `nix-build` on `nix/toolchain.nix` and send the output plus its depedencies to the cache.
To erase the cache:
We don't automate this step at each CI build, as *there is currently no automatic garbage collection of the cache.*
This means we should also monitor the cache's size; if it ever becomes too big we can erase it with:
```
mc rm --recursive --force 'garage/nix/'
@ -157,9 +136,9 @@ nix-shell --run refresh_index
If you want to compile for different architectures, you will need to repeat all these commands for each architecture.
**In practise, and except for debugging, you will never directly run these commands. Release is handled by drone**
**In practice, and except for debugging, you will never directly run these commands. Release is handled by Woodpecker.**
### Drone
### Drone (obsolete)
Our instance is available at [https://drone.deuxfleurs.fr](https://drone.deuxfleurs.fr).
You need an account on [https://git.deuxfleurs.fr](https://git.deuxfleurs.fr) to use it.

View file

@ -0,0 +1,23 @@
+++
title = "Operations & Maintenance"
weight = 50
sort_by = "weight"
template = "documentation.html"
+++
This section contains a number of important information on how to best operate a Garage cluster,
to ensure integrity and availability of your data:
- **[Upgrading Garage](@/documentation/operations/upgrading.md):** General instructions on how to
upgrade your cluster from one version to the next. Instructions specific for each version upgrade
can bef ound in the [working documents](@/documentation/working-documents/_index.md) section.
- **[Layout management](@/documentation/operations/layout.md):** Best practices for using the `garage layout`
commands when adding or removing nodes from your cluster.
- **[Durability and repairs](@/documentation/operations/durability-repairs.md):** How to check for small things
that might be going wrong, and how to recover from such failures.
- **[Recovering from failures](@/documentation/operations/recovering.md):** Garage's first selling point is resilience
to hardware failures. This section explains how to recover from such a failure in the
best possible way.

View file

@ -0,0 +1,147 @@
+++
title = "Durability & Repairs"
weight = 30
+++
To ensure the best durability of your data and to fix any inconsistencies that may
pop up in a distributed system, Garage provides a series of repair operations.
This guide will explain the meaning of each of them and when they should be applied.
# General syntax of repair operations
Repair operations described below are of the form `garage repair <repair_name>`.
These repairs will not launch without the `--yes` flag, which should
be added as follows: `garage repair --yes <repair_name>`.
By default these repair procedures will only run on the Garage node your CLI is
connecting to. To run on all nodes, add the `-a` flag as follows:
`garage repair -a --yes <repair_name>`.
# Data block operations
## Data store scrub {#scrub}
Scrubbing the data store means examining each individual data block to check that
their content is correct, by verifying their hash. Any block found to be corrupted
(e.g. by bitrot or by an accidental manipulation of the datastore) will be
restored from another node that holds a valid copy.
Scrubs are automatically scheduled by Garage to run every 25-35 days (the
actual time is randomized to spread load across nodes). The next scheduled run
can be viewed with `garage worker get`.
A scrub can also be launched manually using `garage repair scrub start`.
To view the status of an ongoing scrub, first find the task ID of the scrub worker
using `garage worker list`. Then, run `garage worker info <scrub_task_id>` to
view detailed runtime statistics of the scrub. To gather cluster-wide information,
this command has to be run on each individual node.
A scrub is a very disk-intensive operation that might slow down your cluster.
You may pause an ongoing scrub using `garage repair scrub pause`, but note that
the scrub will resume automatically 24 hours later as Garage will not let your
cluster run without a regular scrub. If the scrub procedure is too intensive
for your servers and is slowing down your workload, the recommended solution
is to increase the "scrub tranquility" using `garage repair scrub set-tranquility`.
A higher tranquility value will make Garage take longer pauses between two block
verifications. Of course, scrubbing the entire data store will also take longer.
## Block check and resync
In some cases, nodes hold a reference to a block but do not actually have the block
stored on disk. Conversely, they may also have on-disk blocks that are not referenced
any more. To fix both cases, a block repair may be run with `garage repair blocks`.
This will scan the entire block reference counter table to check that the blocks
exist on disk, and will scan the entire disk store to check that stored blocks
are referenced.
It is recommended to run this procedure when changing your cluster layout,
after the metadata tables have finished synchronizing between nodes
(usually a few hours after `garage layout apply`).
## Inspecting lost blocks
In extremely rare situations, data blocks may be unavailable from the entire cluster.
This means that even using `garage repair blocks`, some nodes may be unable
to fetch data blocks for which they hold a reference.
These errors are stored on each node in a list of "block resync errors", i.e.
blocks for which the last resync operation failed.
This list can be inspected using `garage block list-errors`.
These errors usually fall into one of the following categories:
1. a block is still referenced but the object was deleted, this is a case
of metadata reference inconsistency (see below for the fix)
2. a block is referenced by a non-deleted object, but could not be fetched due
to a transient error such as a network failure
3. a block is referenced by a non-deleted object, but could not be fetched due
to a permanent error such as there not being any valid copy of the block on the
entire cluster
To help make the difference between cases 1 and cases 2 and 3, you may use the
`garage block info` command to see which objects hold a reference to each block.
In the second case (transient errors), Garage will try to fetch the block again
after a certain time, so the error should disappear naturally. You can also
request Garage to try to fetch the block immediately using `garage block retry-now`
if you have fixed the transient issue.
If you are confident that you are in the third scenario and that your data block
is definitely lost, then there is no other choice than to declare your S3 objects
as unrecoverable, and to delete them properly from the data store. This can be done
using the `garage block purge` command.
## Rebalancing data directories
In [multi-HDD setups](@/documentation/operations/multi-hdd.md), to ensure that
data blocks are well balanced between storage locations, you may run a
rebalance operation using `garage repair rebalance`. This is useful when
adding storage locations or when capacities of the storage locations have been
changed. Once this is finished, Garage will know for each block of a single
possible location where it can be, which can increase access speed. This
operation will also move out all data from locations marked as read-only.
# Metadata operations
## Metadata snapshotting
It is good practice to setup automatic snapshotting of your metadata database
file, to recover from situations where it becomes corrupted on disk. This can
be done at the filesystem level if you are using ZFS or BTRFS.
Since Garage v0.9.4, Garage is able to take snapshots of the metadata database
itself. This basically amounts to copying the database file, except that it can
be run live while Garage is running without the risk of corruption or
inconsistencies. This can be setup to run automatically on a schedule using
[`metadata_auto_snapshot_interval`](@/documentation/reference-manual/configuration.md#metadata_auto_snapshot_interval).
A snapshot can also be triggered manually using the `garage meta snapshot`
command. Note that taking a snapshot using this method is very intensive as it
requires making a full copy of the database file, so you might prefer using
filesystem-level snapshots if possible. To recover a corrupted node from such a
snapshot, read the instructions
[here](@/documentation/operations/recovering.md#corrupted_meta).
## Metadata table resync
Garage automatically resyncs all entries stored in the metadata tables every hour,
to ensure that all nodes have the most up-to-date version of all the information
they should be holding.
The resync procedure is based on a Merkle tree that allows to efficiently find
differences between nodes.
In some special cases, e.g. before an upgrade, you might want to run a table
resync manually. This can be done using `garage repair tables`.
## Metadata table reference fixes
In some very rare cases where nodes are unavailable, some references between objects
are broken. For instance, if an object is deleted, the underlying versions or data
blocks may still be held by Garage. If you suspect that such corruption has occurred
in your cluster, you can run one of the following repair procedures:
- `garage repair versions`: checks that all versions belong to a non-deleted object, and purges any orphan version
- `garage repair block-refs`: checks that all block references belong to a non-deleted object version, and purges any orphan block reference (this will then allow the blocks to be garbage-collected)
- `garage repair block-rc`: checks that the reference counters for blocks are in sync with the actual number of non-deleted entries in the block reference table

View file

@ -0,0 +1,274 @@
+++
title = "Cluster layout management"
weight = 20
+++
The cluster layout in Garage is a table that assigns to each node a role in
the cluster. The role of a node in Garage can either be a storage node with
a certain capacity, or a gateway node that does not store data and is only
used as an API entry point for faster cluster access.
An introduction to building cluster layouts can be found in the [production deployment](@/documentation/cookbook/real-world.md) page.
In Garage, all of the data that can be stored in a given cluster is divided
into slices which we call *partitions*. Each partition is stored by
one or several nodes in the cluster
(see [`replication_factor`](@/documentation/reference-manual/configuration.md#replication_factor)).
The layout determines the correspondence between these partitions,
which exist on a logical level, and actual storage nodes.
## How cluster layouts work in Garage
A cluster layout is composed of the following components:
- a table of roles assigned to nodes, defined by the user
- an optimal assignation of partitions to nodes, computed by an algorithm that is ran once when calling `garage layout apply` or the ApplyClusterLayout API endpoint
- a version number
Garage nodes will always use the cluster layout with the highest version number.
Garage nodes also maintain and synchronize between them a set of proposed role
changes that haven't yet been applied. These changes will be applied (or
canceled) in the next version of the layout.
All operations on the layout can be realized using the `garage` CLI or using the
[administration API endpoint](@/documentation/reference-manual/admin-api.md).
We give here a description of CLI commands, the admin API semantics are very similar.
The following commands insert modifications to the set of proposed role changes
for the next layout version (but they do not create the new layout immediately):
```bash
garage layout assign [...]
garage layout remove [...]
```
The following command can be used to inspect the layout that is currently set in the cluster
and the changes proposed for the next layout version, if any:
```bash
garage layout show
```
The following commands create a new layout with the specified version number,
that either takes into account the proposed changes or cancels them:
```bash
garage layout apply --version <new_version_number>
garage layout revert --version <new_version_number>
```
The version number of the new layout to create must be 1 + the version number
of the previous layout that existed in the cluster. The `apply` and `revert`
commands will fail otherwise.
## Warnings about Garage cluster layout management
**⚠️ Never make several calls to `garage layout apply` or `garage layout
revert` with the same value of the `--version` flag. Doing so can lead to the
creation of several different layouts with the same version number, in which
case your Garage cluster will become inconsistent until fixed.** If a call to
`garage layout apply` or `garage layout revert` has failed and `garage layout
show` indicates that a new layout with the given version number has not been
set in the cluster, then it is fine to call the command again with the same
version number.
If you are using the `garage` CLI by typing individual commands in your
shell, you shouldn't have much issues as long as you run commands one after
the other and take care of checking the output of `garage layout show`
before applying any changes.
If you are using the `garage` CLI or the admin API to script layout changes,
follow the following recommendations:
- If using the CLI, make all of your `garage` CLI calls to the same RPC host.
If using the admin API, make all of your API calls to the same Garage node. Do
not connect to individual nodes to send them each a piece of the layout changes
you are making, as the changes propagate asynchronously between nodes and might
not all be taken into account at the time when the new layout is applied.
- **Only call `garage layout apply`/ApplyClusterLayout once**, and call it
**strictly after** all of the `layout assign` and `layout remove`
commands/UpdateClusterLayout API calls have returned.
## Understanding unexpected layout calculations
When adding, removing or modifying nodes in a cluster layout, sometimes
unexpected assignations of partitions to node can occur. These assignations
are in fact normal and logical, given the objectives of the algorithm. Indeed,
**the layout algorithm prioritizes moving less data between nodes over
achieving equal distribution of load. It also tries to use all links between
pairs of nodes in equal proportions when moving data.** This section presents
two examples and illustrates how one can control Garage's behavior to obtain
the desired results.
### Example 1
In this example, a cluster is originally composed of 3 nodes in 3 different
zones (data centers). The three nodes are of equal capacity, therefore they
are all fully exploited and all store a copy of all of the data in the cluster.
Then, a fourth node of the same size is added in the datacenter `dc1`.
As illustrated by the following, **Garage will by default not store any data on the new node**:
```
$ garage layout show
==== CURRENT CLUSTER LAYOUT ====
ID Tags Zone Capacity Usable capacity
b10c110e4e854e5a node1 dc1 1000.0 MB 1000.0 MB (100.0%)
a235ac7695e0c54d node2 dc2 1000.0 MB 1000.0 MB (100.0%)
62b218d848e86a64 node3 dc3 1000.0 MB 1000.0 MB (100.0%)
Zone redundancy: maximum
Current cluster layout version: 6
==== STAGED ROLE CHANGES ====
ID Tags Zone Capacity
a11c7cf18af29737 node4 dc1 1000.0 MB
==== NEW CLUSTER LAYOUT AFTER APPLYING CHANGES ====
ID Tags Zone Capacity Usable capacity
b10c110e4e854e5a node1 dc1 1000.0 MB 1000.0 MB (100.0%)
a11c7cf18af29737 node4 dc1 1000.0 MB 0 B (0.0%)
a235ac7695e0c54d node2 dc2 1000.0 MB 1000.0 MB (100.0%)
62b218d848e86a64 node3 dc3 1000.0 MB 1000.0 MB (100.0%)
Zone redundancy: maximum
==== COMPUTATION OF A NEW PARTITION ASSIGNATION ====
Partitions are replicated 3 times on at least 3 distinct zones.
Optimal partition size: 3.9 MB (3.9 MB in previous layout)
Usable capacity / total cluster capacity: 3.0 GB / 4.0 GB (75.0 %)
Effective capacity (replication factor 3): 1000.0 MB
A total of 0 new copies of partitions need to be transferred.
dc1 Tags Partitions Capacity Usable capacity
b10c110e4e854e5a node1 256 (0 new) 1000.0 MB 1000.0 MB (100.0%)
a11c7cf18af29737 node4 0 (0 new) 1000.0 MB 0 B (0.0%)
TOTAL 256 (256 unique) 2.0 GB 1000.0 MB (50.0%)
dc2 Tags Partitions Capacity Usable capacity
a235ac7695e0c54d node2 256 (0 new) 1000.0 MB 1000.0 MB (100.0%)
TOTAL 256 (256 unique) 1000.0 MB 1000.0 MB (100.0%)
dc3 Tags Partitions Capacity Usable capacity
62b218d848e86a64 node3 256 (0 new) 1000.0 MB 1000.0 MB (100.0%)
TOTAL 256 (256 unique) 1000.0 MB 1000.0 MB (100.0%)
```
While unexpected, this is logical because of the following facts:
- storing some data on the new node does not help increase the total quantity
of data that can be stored on the cluster, as the two other zones (`dc2` and
`dc3`) still need to store a full copy of everything, and their capacity is
still the same;
- there is therefore no need to move any data on the new node as this would be pointless;
- moving data to the new node has a cost which the algorithm decides to not pay if not necessary.
This distribution of data can however not be what the administrator wanted: if
they added a new node to `dc1`, it might be because the existing node is too
slow, and they wish to divide its load by half. In that case, what they need to
do to force Garage to distribute the data between the two nodes is to attribute
only half of the capacity to each node in `dc1` (in our example, 500M instead of 1G).
In that case, Garage would determine that to be able to store 1G in total, it
would need to store 500M on the old node and 500M on the added one.
### Example 2
The following example is a slightly different scenario, where `dc1` had two
nodes that were used at 50%, and `dc2` and `dc3` each have one node that is
100% used. All node capacities are the same.
Then, a node from `dc1` is moved into `dc3`. One could expect that the roles of
`dc1` and `dc3` would simply be swapped: the remaining node in `dc1` would be
used at 100%, and the two nodes now in `dc3` would be used at 50%. Instead,
this happens:
```
==== CURRENT CLUSTER LAYOUT ====
ID Tags Zone Capacity Usable capacity
b10c110e4e854e5a node1 dc1 1000.0 MB 500.0 MB (50.0%)
a11c7cf18af29737 node4 dc1 1000.0 MB 500.0 MB (50.0%)
a235ac7695e0c54d node2 dc2 1000.0 MB 1000.0 MB (100.0%)
62b218d848e86a64 node3 dc3 1000.0 MB 1000.0 MB (100.0%)
Zone redundancy: maximum
Current cluster layout version: 8
==== STAGED ROLE CHANGES ====
ID Tags Zone Capacity
a11c7cf18af29737 node4 dc3 1000.0 MB
==== NEW CLUSTER LAYOUT AFTER APPLYING CHANGES ====
ID Tags Zone Capacity Usable capacity
b10c110e4e854e5a node1 dc1 1000.0 MB 1000.0 MB (100.0%)
a235ac7695e0c54d node2 dc2 1000.0 MB 1000.0 MB (100.0%)
62b218d848e86a64 node3 dc3 1000.0 MB 753.9 MB (75.4%)
a11c7cf18af29737 node4 dc3 1000.0 MB 246.1 MB (24.6%)
Zone redundancy: maximum
==== COMPUTATION OF A NEW PARTITION ASSIGNATION ====
Partitions are replicated 3 times on at least 3 distinct zones.
Optimal partition size: 3.9 MB (3.9 MB in previous layout)
Usable capacity / total cluster capacity: 3.0 GB / 4.0 GB (75.0 %)
Effective capacity (replication factor 3): 1000.0 MB
A total of 128 new copies of partitions need to be transferred.
dc1 Tags Partitions Capacity Usable capacity
b10c110e4e854e5a node1 256 (128 new) 1000.0 MB 1000.0 MB (100.0%)
TOTAL 256 (256 unique) 1000.0 MB 1000.0 MB (100.0%)
dc2 Tags Partitions Capacity Usable capacity
a235ac7695e0c54d node2 256 (0 new) 1000.0 MB 1000.0 MB (100.0%)
TOTAL 256 (256 unique) 1000.0 MB 1000.0 MB (100.0%)
dc3 Tags Partitions Capacity Usable capacity
62b218d848e86a64 node3 193 (0 new) 1000.0 MB 753.9 MB (75.4%)
a11c7cf18af29737 node4 63 (0 new) 1000.0 MB 246.1 MB (24.6%)
TOTAL 256 (256 unique) 2.0 GB 1000.0 MB (50.0%)
```
As we can see, the node that was moved to `dc3` (node4) is only used at 25% (approximatively),
whereas the node that was already in `dc3` (node3) is used at 75%.
This can be explained by the following:
- node1 will now be the only node remaining in `dc1`, thus it has to store all
of the data in the cluster. Since it was storing only half of it before, it has
to retrieve the other half from other nodes in the cluster.
- The data which it does not have is entirely stored by the other node that was
in `dc1` and that is now in `dc3` (node4). There is also a copy of it on node2
and node3 since both these nodes have a copy of everything.
- node3 and node4 are the two nodes that will now be in a datacenter that is
under-utilized (`dc3`), this means that those are the two candidates from which
data can be removed to be moved to node1.
- Garage will move data in equal proportions from all possible sources, in this
case it means that it will tranfer 25% of the entire data set from node3 to
node1 and another 25% from node4 to node1.
This explains why node3 ends with 75% utilization (100% from before minus 25%
that is moved to node1), and node4 ends with 25% (50% from before minus 25%
that is moved to node1).
This illustrates the second principle of the layout computation: **if there is
a choice in moving data out of some nodes, then all links between pairs of
nodes are used in equal proportions** (this is approximately true, there is
randomness in the algorithm to achieve this so there might be some small
fluctuations, as we see above).

View file

@ -0,0 +1,101 @@
+++
title = "Multi-HDD support"
weight = 15
+++
Since v0.9, Garage natively supports nodes that have several storage drives
for storing data blocks (not for metadata storage).
## Initial setup
To set up a new Garage storage node with multiple HDDs,
format and mount all your drives in different directories,
and use a Garage configuration as follows:
```toml
data_dir = [
{ path = "/path/to/hdd1", capacity = "2T" },
{ path = "/path/to/hdd2", capacity = "4T" },
]
```
Garage will automatically balance all blocks stored by the node
among the different specified directories, proportionnally to the
specified capacities.
## Updating the list of storage locations
If you add new storage locations to your `data_dir`,
Garage will not rebalance existing data between storage locations.
Newly written blocks will be balanced proportionnally to the specified capacities,
and existing data may be moved between drives to improve balancing,
but only opportunistically when a data block is re-written (e.g. an object
is re-uploaded, or an object with a duplicate block is uploaded).
To understand precisely what is happening, we need to dive in to how Garage
splits data among the different storage locations.
First of all, Garage divides the set of all possible block hashes
in a fixed number of slices (currently 1024), and assigns
to each slice a primary storage location among the specified data directories.
The number of slices having their primary location in each data directory
is proportionnal to the capacity specified in the config file.
When Garage receives a block to write, it will always write it in the primary
directory of the slice that contains its hash.
Now, to be able to not lose existing data blocks when storage locations
are added, Garage also keeps a list of secondary data directories
for all of the hash slices. Secondary data directories for a slice indicates
storage locations that once were primary directories for that slice, i.e. where
Garage knows that data blocks of that slice might be stored.
When Garage is requested to read a certain data block,
it will first look in the primary storage directory of its slice,
and if it doesn't find it there it goes through all of the secondary storage
locations until it finds it. This allows Garage to continue operating
normally when storage locations are added, without having to shuffle
files between drives to place them in the correct location.
This relatively simple strategy works well but does not ensure that data
is correctly balanced among drives according to their capacity.
To rebalance data, two strategies can be used:
- Lazy rebalancing: when a block is re-written (e.g. the object is re-uploaded),
Garage checks whether the existing copy is in the primary directory of the slice
or in a secondary directory. If the current copy is in a secondary directory,
Garage re-writes a copy in the primary directory and deletes the one from the
secondary directory. This might never end up rebalancing everything if there
are data blocks that are only read and never written.
- Active rebalancing: an operator of a Garage node can explicitly launch a repair
procedure that rebalances the data directories, moving all blocks to their
primary location. Once done, all secondary locations for all hash slices are
removed so that they won't be checked anymore when looking for a data block.
## Read-only storage locations
If you would like to move all data blocks from an existing data directory to one
or several new data directories, mark the old directory as read-only:
```toml
data_dir = [
{ path = "/path/to/old_data", read_only = true },
{ path = "/path/to/new_hdd1", capacity = "2T" },
{ path = "/path/to/new_hdd2", capacity = "4T" },
]
```
Garage will be able to read requested blocks from the read-only directory.
Garage will also move data out of the read-only directory either progressively
(lazy rebalancing) or if requested explicitly (active rebalancing).
Once an active rebalancing has finished, your read-only directory should be empty:
it might still contain subdirectories, but no data files. You can check that
it contains no files using:
```bash
find -type f /path/to/old_data # should not print anything
```
at which point it can be removed from the `data_dir` list in your config file.

View file

@ -1,11 +1,11 @@
+++
title = "Recovering from failures"
weight = 50
weight = 40
+++
Garage is meant to work on old, second-hand hardware.
In particular, this makes it likely that some of your drives will fail, and some manual intervention will be needed.
Fear not! For Garage is fully equipped to handle drive failures, in most common cases.
Fear not! Garage is fully equipped to handle drive failures, in most common cases.
## A note on availability of Garage
@ -108,3 +108,57 @@ garage layout apply # once satisfied, apply the changes
Garage will then start synchronizing all required data on the new node.
This process can be monitored using the `garage stats -a` command.
## Replacement scenario 3: corrupted metadata {#corrupted_meta}
In some cases, your metadata DB file might become corrupted, for instance if
your node suffered a power outage and did not shut down properly. In this case,
you can recover without having to change the node ID and rebuilding a cluster
layout. This means that data blocks will not need to be shuffled around, you
must simply find a way to repair the metadata file. The best way is generally
to discard the corrupted file and recover it from another source.
First of all, start by locating the database file in your metadata directory,
which [depends on your `db_engine`
choice](@/documentation/reference-manual/configuration.md#db_engine). Then,
your recovery options are as follows:
- **Option 1: resyncing from other nodes.** In case your cluster is replicated
with two or three copies, you can simply delete the database file, and Garage
will resync from other nodes. To do so, stop Garage, delete the database file
or directory, and restart Garage. Then, do a full table repair by calling
`garage repair -a --yes tables`. This will take a bit of time to complete as
the new node will need to receive copies of the metadata tables from the
network.
- **Option 2: restoring a snapshot taken by Garage.** Since v0.9.4, Garage can
[automatically take regular
snapshots](@/documentation/reference-manual/configuration.md#metadata_auto_snapshot_interval)
of your metadata DB file. This file or directory should be located under
`<metadata_dir>/snapshots`, and is named according to the UTC time at which it
was taken. Stop Garage, discard the database file/directory and replace it by the
snapshot you want to use. For instance, in the case of LMDB:
```bash
cd $METADATA_DIR
mv db.lmdb db.lmdb.bak
cp -r snapshots/2024-03-15T12:13:52Z db.lmdb
```
And for Sqlite:
```bash
cd $METADATA_DIR
mv db.sqlite db.sqlite.bak
cp snapshots/2024-03-15T12:13:52Z db.sqlite
```
Then, restart Garage and run a full table repair by calling `garage repair -a
--yes tables`. This should run relatively fast as only the changes that
occurred since the snapshot was taken will need to be resynchronized. Of
course, if your cluster is not replicated, you will lose all changes that
occurred since the snapshot was taken.
- **Option 3: restoring a filesystem-level snapshot.** If you are using ZFS or
BTRFS to snapshot your metadata partition, refer to their specific
documentation on rolling back or copying files from an old snapshot.

View file

@ -1,6 +1,6 @@
+++
title = "Upgrading Garage"
weight = 60
weight = 10
+++
Garage is a stateful clustered application, where all nodes are communicating together and share data structures.
@ -9,7 +9,7 @@ On a new version release, there is 2 possibilities:
- protocols and data structures remained the same ➡️ this is a **minor upgrade**
- protocols or data structures changed ➡️ this is a **major upgrade**
You can quickly now what type of update you will have to operate by looking at the version identifier:
You can quickly know what type of update you will have to operate by looking at the version identifier:
when we require our users to do a major upgrade, we will always bump the first nonzero component of the version identifier
(e.g. from v0.7.2 to v0.8.0).
Conversely, for versions that only require a minor upgrade, the first nonzero component will always stay the same (e.g. from v0.8.0 to v0.8.1).
@ -58,7 +58,7 @@ From a high level perspective, a major upgrade looks like this:
### Major upgarades with minimal downtime
There is only one operation that has to be coordinated cluster-wide: the passage of one version of the internal RPC protocol to the next.
There is only one operation that has to be coordinated cluster-wide: the switch of one version of the internal RPC protocol to the next.
This means that an upgrade with very limited downtime can simply be performed from one major version to the next by restarting all nodes
simultaneously in the new version.
The downtime will simply be the time required for all nodes to stop and start again, which should be less than a minute.
@ -73,6 +73,18 @@ The entire procedure would look something like this:
You can do all of the nodes in a single zone at once as that won't impact global cluster availability.
Do not try to make a backup of the metadata folder of a running node.
**Since Garage v0.9.4,** you can use the `garage meta snapshot --all` command
to take a simultaneous snapshot of the metadata database files of all your
nodes. This avoids the tedious process of having to take them down one by
one before upgrading. Be careful that if automatic snapshotting is enabled,
Garage only keeps the last two snapshots and deletes older ones, so you might
want to disable automatic snapshotting in your upgraded configuration file
until you have confirmed that the upgrade ran successfully. In addition to
snapshotting the metadata databases of your nodes, you should back-up at
least the `cluster_layout` file of one of your Garage instances (this file
should be the same on all nodes and you can copy it safely while Garage is
running).
3. Prepare your binaries and configuration files for the new Garage version
4. Restart all nodes simultaneously in the new version
@ -80,6 +92,6 @@ The entire procedure would look something like this:
5. If any specific migration procedure is required, it is usually in one of the two cases:
- It can be run on online nodes after the new version has started, during regular cluster operation.
- it has to be run offline
- it has to be run offline, in which case you will have to again take all nodes offline one after the other to run the repair
For this last step, please refer to the specific documentation pertaining to the version upgrade you are doing.

View file

@ -1,6 +1,6 @@
+++
title = "Quick Start"
weight = 0
weight = 10
sort_by = "weight"
template = "documentation.html"
+++
@ -35,10 +35,20 @@ Place this binary somewhere in your `$PATH` so that you can invoke the `garage`
command directly (for instance you can copy the binary in `/usr/local/bin`
or in `~/.local/bin`).
You may also check whether your distribution already includes a
[binary package for Garage](@/documentation/cookbook/binary-packages.md).
If a binary of the last version is not available for your architecture,
or if you want a build customized for your system,
you can [build Garage from source](@/documentation/cookbook/from-source.md).
If none of these option work for you, you can also run Garage in a Docker
container. When using Docker, the commands used in this guide will not work
anymore. We recommend reading the tutorial on [configuring a
multi-node cluster](@/documentation/cookbook/real-world.md) to learn about
using Garage as a Docker container. For simplicity, a minimal command to launch
Garage using Docker is provided in this quick start guide as well.
## Configuring and starting Garage
@ -54,9 +64,9 @@ to generate unique and private secrets for security reasons:
cat > garage.toml <<EOF
metadata_dir = "/tmp/meta"
data_dir = "/tmp/data"
db_engine = "lmdb"
db_engine = "sqlite"
replication_mode = "none"
replication_factor = 1
rpc_bind_addr = "[::]:3901"
rpc_public_addr = "127.0.0.1:3901"
@ -76,14 +86,17 @@ index = "index.html"
api_bind_addr = "[::]:3904"
[admin]
api_bind_addr = "0.0.0.0:3903"
api_bind_addr = "[::]:3903"
admin_token = "$(openssl rand -base64 32)"
metrics_token = "$(openssl rand -base64 32)"
EOF
```
Now that your configuration file has been created, you can put
it in the right place. By default, garage looks at **`/etc/garage.toml`.**
See the [Configuration file format](https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/)
for complete options and values.
Now that your configuration file has been created, you may save it to the directory of your choice.
By default, Garage looks for **`/etc/garage.toml`.**
You can also store it somewhere else, but you will have to specify `-c path/to/garage.toml`
at each invocation of the `garage` binary (for example: `garage -c ./garage.toml server`, `garage -c ./garage.toml status`).
@ -100,16 +113,39 @@ your data to be persisted properly.
### Launching the Garage server
Use the following command to launch the Garage server with our configuration file:
Use the following command to launch the Garage server:
```
garage server
garage -c path/to/garage.toml server
```
You can tune Garage's verbosity as follows (from less verbose to more verbose):
If you have placed the `garage.toml` file in `/etc` (its default location), you can simply run `garage server`.
Alternatively, if you cannot or do not wish to run the Garage binary directly,
you may use Docker to run Garage in a container using the following command:
```bash
docker run \
-d \
--name garaged \
-p 3900:3900 -p 3901:3901 -p 3902:3902 -p 3903:3903 \
-v /etc/garage.toml:/path/to/garage.toml \
-v /var/lib/garage/meta:/path/to/garage/meta \
-v /var/lib/garage/data:/path/to/garage/data \
dxflrs/garage:v0.9.4
```
RUST_LOG=garage=info garage server
Under Linux, you can substitute `--network host` for `-p 3900:3900 -p 3901:3901 -p 3902:3902 -p 3903:3903`
#### Troubleshooting
Ensure your configuration file, `metadata_dir` and `data_dir` are readable by the user running the `garage` server or Docker.
You can tune Garage's verbosity by setting the `RUST_LOG=` environment variable. \
Available log levels are (from less verbose to more verbose): `error`, `warn`, `info` *(default)*, `debug` and `trace`.
```bash
RUST_LOG=garage=info garage server # default
RUST_LOG=garage=debug garage server
RUST_LOG=garage=trace garage server
```
@ -123,7 +159,10 @@ Log level `debug` can help you check why your S3 API calls are not working.
The `garage` utility is also used as a CLI tool to configure your Garage deployment.
It uses values from the TOML configuration file to find the Garage daemon running on the
local node, therefore if your configuration file is not at `/etc/garage.toml` you will
again have to specify `-c path/to/garage.toml`.
again have to specify `-c path/to/garage.toml` at each invocation.
If you are running Garage in a Docker container, you can set `alias garage="docker exec -ti <container name> /garage"`
to use the Garage binary inside your container.
If the `garage` CLI is able to correctly detect the parameters of your local Garage node,
the following command should be enough to show the status of your cluster:
@ -137,7 +176,7 @@ This should show something like this:
```
==== HEALTHY NODES ====
ID Hostname Address Tag Zone Capacity
563e1ac825ee3323 linuxbox 127.0.0.1:3901 NO ROLE ASSIGNED
563e1ac825ee3323 linuxbox 127.0.0.1:3901 NO ROLE ASSIGNED
```
## Creating a cluster layout
@ -150,12 +189,12 @@ For our test deployment, we are using only one node. The way in which we configu
it does not matter, you can simply write:
```bash
garage layout assign -z dc1 -c 1 <node_id>
garage layout assign -z dc1 -c 1G <node_id>
```
where `<node_id>` corresponds to the identifier of the node shown by `garage status` (first column).
You can enter simply a prefix of that identifier.
For instance here you could write just `garage layout assign -z dc1 -c 1 563e`.
For instance here you could write just `garage layout assign -z dc1 -c 1G 563e`.
The layout then has to be applied to the cluster, using:
@ -206,7 +245,7 @@ one key can access multiple buckets, multiple keys can access one bucket.
Create an API key using the following command:
```
garage key new --name nextcloud-app-key
garage key create nextcloud-app-key
```
The output should look as follows:
@ -245,7 +284,7 @@ garage bucket info nextcloud-bucket
```
## Uploading and downlading from Garage
## Uploading and downloading from Garage
To download and upload files on garage, we can use a third-party tool named `awscli`.
@ -266,12 +305,14 @@ named `~/.awsrc` with this content:
export AWS_ACCESS_KEY_ID=xxxx # put your Key ID here
export AWS_SECRET_ACCESS_KEY=xxxx # put your Secret key here
export AWS_DEFAULT_REGION='garage'
export AWS_ENDPOINT='http://localhost:3900'
export AWS_ENDPOINT_URL='http://localhost:3900'
function aws { command aws --endpoint-url $AWS_ENDPOINT $@ ; }
aws --version
```
Note you need to have at least `awscli` `>=1.29.0` or `>=2.13.0`, otherwise you
need to specify `--endpoint-url` explicitly on each `awscli` invocation.
Now, each time you want to use `awscli` on this target, run:
```bash

View file

@ -1,6 +1,6 @@
+++
title = "Reference Manual"
weight = 5
weight = 60
sort_by = "weight"
template = "documentation.html"
+++

View file

@ -8,18 +8,21 @@ listen address is specified in the `[admin]` section of the configuration
file (see [configuration file
reference](@/documentation/reference-manual/configuration.md))
**WARNING.** At this point, there is no comittement to stability of the APIs described in this document.
We will bump the version numbers prefixed to each API endpoint at each time the syntax
**WARNING.** At this point, there is no commitment to the stability of the APIs described in this document.
We will bump the version numbers prefixed to each API endpoint each time the syntax
or semantics change, meaning that code that relies on these endpoint will break
when changes are introduced.
The Garage administration API was introduced in version 0.7.2, this document
does not apply to older versions of Garage.
Versions:
- Before Garage 0.7.2 - no admin API
- Garage 0.7.2 - admin APIv0
- Garage 0.9.0 - admin APIv1, deprecate admin APIv0
## Access control
The admin API uses two different tokens for acces control, that are specified in the config file's `[admin]` section:
The admin API uses two different tokens for access control, that are specified in the config file's `[admin]` section:
- `metrics_token`: the token for accessing the Metrics endpoint (if this token
is not set in the config file, the Metrics endpoint can be accessed without
@ -39,15 +42,101 @@ Authorization: Bearer <token>
## Administration API endpoints
### Metrics-related endpoints
#### Metrics `GET /metrics`
### Metrics `GET /metrics`
Returns internal Garage metrics in Prometheus format.
The metrics are directly documented when returned by the API.
**Example:**
```
$ curl -i http://localhost:3903/metrics
HTTP/1.1 200 OK
content-type: text/plain; version=0.0.4
content-length: 12145
date: Tue, 08 Aug 2023 07:25:05 GMT
# HELP api_admin_error_counter Number of API calls to the various Admin API endpoints that resulted in errors
# TYPE api_admin_error_counter counter
api_admin_error_counter{api_endpoint="CheckWebsiteEnabled",status_code="400"} 1
api_admin_error_counter{api_endpoint="CheckWebsiteEnabled",status_code="404"} 3
# HELP api_admin_request_counter Number of API calls to the various Admin API endpoints
# TYPE api_admin_request_counter counter
api_admin_request_counter{api_endpoint="CheckWebsiteEnabled"} 7
api_admin_request_counter{api_endpoint="Health"} 3
# HELP api_admin_request_duration Duration of API calls to the various Admin API endpoints
...
```
### Health `GET /health`
Returns `200 OK` if enough nodes are up to have a quorum (ie. serve requests),
otherwise returns `503 Service Unavailable`.
**Example:**
```
$ curl -i http://localhost:3903/health
HTTP/1.1 200 OK
content-type: text/plain
content-length: 102
date: Tue, 08 Aug 2023 07:22:38 GMT
Garage is fully operational
Consult the full health check API endpoint at /v0/health for more details
```
### On-demand TLS `GET /check`
To prevent abuse for on-demand TLS, Caddy developers have specified an endpoint that can be queried by the reverse proxy
to know if a given domain is allowed to get a certificate. Garage implements these endpoints to tell if a given domain is handled by Garage or is garbage.
Garage responds with the following logic:
- If the domain matches the pattern `<bucket-name>.<s3_api.root_domain>`, returns 200 OK
- If the domain matches the pattern `<bucket-name>.<s3_web.root_domain>` and website is configured for `<bucket>`, returns 200 OK
- If the domain matches the pattern `<bucket-name>` and website is configured for `<bucket>`, returns 200 OK
- Otherwise, returns 404 Not Found, 400 Bad Request or 5xx requests.
*Note 1: because in the path-style URL mode, there is only one domain that is not known by Garage, hence it is not supported by this API endpoint.
You must manually declare the domain in your reverse-proxy. Idem for K2V.*
*Note 2: buckets in a user's namespace are not supported yet by this endpoint. This is a limitation of this endpoint currently.*
**Example:** Suppose a Garage instance is configured with `s3_api.root_domain = .s3.garage.localhost` and `s3_web.root_domain = .web.garage.localhost`.
With a private `media` bucket (name in the global namespace, website is disabled), the endpoint will feature the following behavior:
```
$ curl -so /dev/null -w "%{http_code}" http://localhost:3903/check?domain=media.s3.garage.localhost
200
$ curl -so /dev/null -w "%{http_code}" http://localhost:3903/check?domain=media
400
$ curl -so /dev/null -w "%{http_code}" http://localhost:3903/check?domain=media.web.garage.localhost
400
```
With a public `example.com` bucket (name in the global namespace, website is activated), the endpoint will feature the following behavior:
```
$ curl -so /dev/null -w "%{http_code}" http://localhost:3903/check?domain=example.com.s3.garage.localhost
200
$ curl -so /dev/null -w "%{http_code}" http://localhost:3903/check?domain=example.com
200
$ curl -so /dev/null -w "%{http_code}" http://localhost:3903/check?domain=example.com.web.garage.localhost
200
```
**References:**
- [Using On-Demand TLS](https://caddyserver.com/docs/automatic-https#using-on-demand-tls)
- [Add option for a backend check to approve use of on-demand TLS](https://github.com/caddyserver/caddy/pull/1939)
- [Serving tens of thousands of domains over HTTPS with Caddy](https://caddy.community/t/serving-tens-of-thousands-of-domains-over-https-with-caddy/11179)
### Cluster operations
These endpoints are defined on a dedicated [Redocly page](https://garagehq.deuxfleurs.fr/api/garage-admin-v0.html). You can also download its [OpenAPI specification](https://garagehq.deuxfleurs.fr/api/garage-admin-v0.yml).
These endpoints have a dedicated OpenAPI spec.
- APIv1 - [HTML spec](https://garagehq.deuxfleurs.fr/api/garage-admin-v1.html) - [OpenAPI YAML](https://garagehq.deuxfleurs.fr/api/garage-admin-v1.yml)
- APIv0 (deprecated) - [HTML spec](https://garagehq.deuxfleurs.fr/api/garage-admin-v0.html) - [OpenAPI YAML](https://garagehq.deuxfleurs.fr/api/garage-admin-v0.yml)
Requesting the API from the command line can be as simple as running:

View file

@ -8,39 +8,57 @@ weight = 20
Here is an example `garage.toml` configuration file that illustrates all of the possible options:
```toml
replication_factor = 3
consistency_mode = "consistent"
metadata_dir = "/var/lib/garage/meta"
data_dir = "/var/lib/garage/data"
metadata_fsync = true
data_fsync = false
disable_scrub = false
use_local_tz = false
metadata_auto_snapshot_interval = "6h"
db_engine = "lmdb"
block_size = 1048576
block_size = "1M"
block_ram_buffer_max = "256MiB"
sled_cache_capacity = 134217728
sled_flush_every_ms = 2000
replication_mode = "3"
lmdb_map_size = "1T"
compression_level = 1
rpc_secret = "4425f5c26c5e11581d3223904324dcb5b5d5dfb14e5e7f35e38c595424f5f1e6"
rpc_bind_addr = "[::]:3901"
rpc_bind_outgoing = false
rpc_public_addr = "[fc00:1::1]:3901"
# or set rpc_public_adr_subnet to filter down autodiscovery to a subnet:
# rpc_public_addr_subnet = "2001:0db8:f00:b00:/64"
allow_world_readable_secrets = false
bootstrap_peers = [
"563e1ac825ee3323aa441e72c26d1030d6d4414aeb3dd25287c531e7fc2bc95d@[fc00:1::1]:3901",
"86f0f26ae4afbd59aaf9cfb059eefac844951efd5b8caeec0d53f4ed6c85f332[fc00:1::2]:3901",
"86f0f26ae4afbd59aaf9cfb059eefac844951efd5b8caeec0d53f4ed6c85f332@[fc00:1::2]:3901",
"681456ab91350f92242e80a531a3ec9392cb7c974f72640112f90a600d7921a4@[fc00:B::1]:3901",
"212fd62eeaca72c122b45a7f4fa0f55e012aa5e24ac384a72a3016413fa724ff@[fc00:F::1]:3901",
]
[consul_discovery]
api = "catalog"
consul_http_addr = "http://127.0.0.1:8500"
service_name = "garage-daemon"
ca_cert = "/etc/consul/consul-ca.crt"
client_cert = "/etc/consul/consul-client.crt"
client_key = "/etc/consul/consul-key.crt"
# for `agent` API mode, unset client_cert and client_key, and optionally enable `token`
# token = "abcdef-01234-56789"
tls_skip_verify = false
tags = [ "dns-enabled" ]
meta = { dns-acl = "allow trusted" }
[kubernetes_discovery]
namespace = "garage"
@ -59,8 +77,8 @@ root_domain = ".web.garage"
[admin]
api_bind_addr = "0.0.0.0:3903"
metrics_token = "cacce0b2de4bc2d9f5b5fdff551e01ac1496055aed248202d415398987e35f81"
admin_token = "ae8cb40ea7368bbdbb6430af11cca7da833d3458a5f52086f4e805a570fb5c2a"
metrics_token = "BCAdFjoa9G0KJR0WXnHHm7fs1ZAbfpI8iIZ+Z/a2NgI="
admin_token = "UkLeGWEvHnXBqnueR3ISEMWpOnm40jH2tM2HnnL/0F4="
trace_sink = "http://localhost:4317"
```
@ -68,97 +86,91 @@ The following gives details about each available configuration option.
## Available configuration options
### `metadata_dir`
### Index
The directory in which Garage will store its metadata. This contains the node identifier,
the network configuration and the peer list, the list of buckets and keys as well
as the index of all objects, object version and object blocks.
[Environment variables](#env_variables).
Store this folder on a fast SSD drive if possible to maximize Garage's performance.
Top-level configuration options:
[`allow_world_readable_secrets`](#allow_world_readable_secrets),
[`block_ram_buffer_max`](#block_ram_buffer_max),
[`block_size`](#block_size),
[`bootstrap_peers`](#bootstrap_peers),
[`compression_level`](#compression_level),
[`data_dir`](#data_dir),
[`data_fsync`](#data_fsync),
[`db_engine`](#db_engine),
[`disable_scrub`](#disable_scrub),
[`use_local_tz`](#use_local_tz),
[`lmdb_map_size`](#lmdb_map_size),
[`metadata_auto_snapshot_interval`](#metadata_auto_snapshot_interval),
[`metadata_dir`](#metadata_dir),
[`metadata_fsync`](#metadata_fsync),
[`replication_factor`](#replication_factor),
[`consistency_mode`](#consistency_mode),
[`rpc_bind_addr`](#rpc_bind_addr),
[`rpc_bind_outgoing`](#rpc_bind_outgoing),
[`rpc_public_addr`](#rpc_public_addr),
[`rpc_public_addr_subnet`](#rpc_public_addr_subnet)
[`rpc_secret`/`rpc_secret_file`](#rpc_secret).
### `data_dir`
The `[consul_discovery]` section:
[`api`](#consul_api),
[`ca_cert`](#consul_ca_cert),
[`client_cert`](#consul_client_cert_and_key),
[`client_key`](#consul_client_cert_and_key),
[`consul_http_addr`](#consul_http_addr),
[`meta`](#consul_tags_and_meta),
[`service_name`](#consul_service_name),
[`tags`](#consul_tags_and_meta),
[`tls_skip_verify`](#consul_tls_skip_verify),
[`token`](#consul_token).
The directory in which Garage will store the data blocks of objects.
This folder can be placed on an HDD. The space available for `data_dir`
should be counted to determine a node's capacity
when [adding it to the cluster layout](@/documentation/cookbook/real-world.md).
The `[kubernetes_discovery]` section:
[`namespace`](#kube_namespace),
[`service_name`](#kube_service_name),
[`skip_crd`](#kube_skip_crd).
### `db_engine` (since `v0.8.0`)
The `[s3_api]` section:
[`api_bind_addr`](#s3_api_bind_addr),
[`root_domain`](#s3_root_domain),
[`s3_region`](#s3_region).
By default, Garage uses the Sled embedded database library
to store its metadata on-disk. Since `v0.8.0`, Garage can use alternative storage backends as follows:
The `[s3_web]` section:
[`bind_addr`](#web_bind_addr),
[`root_domain`](#web_root_domain).
| DB engine | `db_engine` value | Database path |
| --------- | ----------------- | ------------- |
| [Sled](https://sled.rs) | `"sled"` | `<metadata_dir>/db/` |
| [LMDB](https://www.lmdb.tech) | `"lmdb"` | `<metadata_dir>/db.lmdb/` |
| [Sqlite](https://sqlite.org) | `"sqlite"` | `<metadata_dir>/db.sqlite` |
The `[admin]` section:
[`api_bind_addr`](#admin_api_bind_addr),
[`metrics_token`/`metrics_token_file`](#admin_metrics_token),
[`admin_token`/`admin_token_file`](#admin_token),
[`trace_sink`](#admin_trace_sink),
Performance characteristics of the different DB engines are as follows:
### Environment variables {#env_variables}
- Sled: the default database engine, which tends to produce
large data files and also has performance issues, especially when the metadata folder
is on a traditional HDD and not on SSD.
- LMDB: the recommended alternative on 64-bit systems,
much more space-efficiant and slightly faster. Note that the data format of LMDB is not portable
between architectures, so for instance the Garage database of an x86-64
node cannot be moved to an ARM64 node. Also note that, while LMDB can technically be used on 32-bit systems,
this will limit your node to very small database sizes due to how LMDB works; it is therefore not recommended.
- Sqlite: Garage supports Sqlite as a storage backend for metadata,
however it may have issues and is also very slow in its current implementation,
so it is not recommended to be used for now.
The following configuration parameter must be specified as an environment
variable, it does not exist in the configuration file:
It is possible to convert Garage's metadata directory from one format to another with a small utility named `convert_db`,
which can be downloaded at the following locations:
[for amd64](https://garagehq.deuxfleurs.fr/_releases/convert_db/amd64/convert_db),
[for i386](https://garagehq.deuxfleurs.fr/_releases/convert_db/i386/convert_db),
[for arm64](https://garagehq.deuxfleurs.fr/_releases/convert_db/arm64/convert_db),
[for arm](https://garagehq.deuxfleurs.fr/_releases/convert_db/arm/convert_db).
The `convert_db` utility is used as folows:
- `GARAGE_LOG_TO_SYSLOG` (since v0.9.4): set this to `1` or `true` to make the
Garage daemon send its logs to `syslog` (using the libc `syslog` function)
instead of printing to stderr.
```
convert-db -a <input db engine> -i <input db path> \
-b <output db engine> -o <output db path>
```
The following environment variables can be used to override the corresponding
values in the configuration file:
Make sure to specify the full database path as presented in the table above,
and not just the path to the metadata directory.
- [`GARAGE_ALLOW_WORLD_READABLE_SECRETS`](#allow_world_readable_secrets)
- [`GARAGE_RPC_SECRET` and `GARAGE_RPC_SECRET_FILE`](#rpc_secret)
- [`GARAGE_ADMIN_TOKEN` and `GARAGE_ADMIN_TOKEN_FILE`](#admin_token)
- [`GARAGE_METRICS_TOKEN` and `GARAGE_METRICS_TOKEN`](#admin_metrics_token)
### `block_size`
Garage splits stored objects in consecutive chunks of size `block_size`
(except the last one which might be smaller). The default size is 1MB and
should work in most cases. We recommend increasing it to e.g. 10MB if
you are using Garage to store large files and have fast network connections
between all nodes (e.g. 1gbps).
### Top-level configuration options
If you are interested in tuning this, feel free to do so (and remember to
report your findings to us!). When this value is changed for a running Garage
installation, only files newly uploaded will be affected. Previously uploaded
files will remain available. This however means that chunks from existing files
will not be deduplicated with chunks from newly uploaded files, meaning you
might use more storage space that is optimally possible.
#### `replication_factor` {#replication_factor}
### `sled_cache_capacity`
The replication factor can be any positive integer smaller or equal the node count in your cluster.
The chosen replication factor has a big impact on the cluster's failure tolerancy and performance characteristics.
This parameter can be used to tune the capacity of the cache used by
[sled](https://sled.rs), the database Garage uses internally to store metadata.
Tune this to fit the RAM you wish to make available to your Garage instance.
This value has a conservative default (128MB) so that Garage doesn't use too much
RAM by default, but feel free to increase this for higher performance.
### `sled_flush_every_ms`
This parameters can be used to tune the flushing interval of sled.
Increase this if sled is thrashing your SSD, at the risk of losing more data in case
of a power outage (though this should not matter much as data is replicated on other
nodes). The default value, 2000ms, should be appropriate for most use cases.
### `replication_mode`
Garage supports the following replication modes:
- `none` or `1`: data stored on Garage is stored on a single node. There is no
- `1`: data stored on Garage is stored on a single node. There is no
redundancy, and data will be unavailable as soon as one node fails or its
network is disconnected. Do not use this for anything else than test
deployments.
@ -169,17 +181,6 @@ Garage supports the following replication modes:
before losing data. Data remains available in read-only mode when one node is
down, but write operations will fail.
- `2-dangerous`: a variant of mode `2`, where written objects are written to
the second replica asynchronously. This means that Garage will return `200
OK` to a PutObject request before the second copy is fully written (or even
before it even starts being written). This means that data can more easily
be lost if the node crashes before a second copy can be completed. This
also means that written objects might not be visible immediately in read
operations. In other words, this mode severely breaks the consistency and
durability guarantees of standard Garage cluster operation. Benefits of
this mode: you can still write to your cluster when one node is
unavailable.
- `3`: data stored on Garage will be stored on three different nodes, if
possible each in a different zones. Garage tolerates two node failure, or
several node failures but in no more than two zones (in a deployment with at
@ -187,57 +188,309 @@ Garage supports the following replication modes:
or node failures are only in a single zone, reading and writing data to
Garage can continue normally.
- `3-degraded`: a variant of replication mode `3`, that lowers the read
quorum to `1`, to allow you to read data from your cluster when several
nodes (or nodes in several zones) are unavailable. In this mode, Garage
does not provide read-after-write consistency anymore. The write quorum is
still 2, ensuring that data successfully written to Garage is stored on at
least two nodes.
- `3-dangerous`: a variant of replication mode `3` that lowers both the read
and write quorums to `1`, to allow you to both read and write to your
cluster when several nodes (or nodes in several zones) are unavailable. It
is the least consistent mode of operation proposed by Garage, and also one
that should probably never be used.
- `5`, `7`, ...: When setting the replication factor above 3, it is most useful to
choose an uneven value, since for every two copies added, one more node can fail
before losing the ability to write and read to the cluster.
Note that in modes `2` and `3`,
if at least the same number of zones are available, an arbitrary number of failures in
any given zone is tolerated as copies of data will be spread over several zones.
**Make sure `replication_mode` is the same in the configuration files of all nodes.
**Make sure `replication_factor` is the same in the configuration files of all nodes.
Never run a Garage cluster where that is not the case.**
It is technically possible to change the replication factor although it's a
dangerous operation that is not officially supported. This requires you to
delete the existing cluster layout and create a new layout from scratch,
meaning that a full rebalancing of your cluster's data will be needed. To do
it, shut down your cluster entirely, delete the `custer_layout` files in the
meta directories of all your nodes, update all your configuration files with
the new `replication_factor` parameter, restart your cluster, and then create a
new layout with all the nodes you want to keep. Rebalancing data will take
some time, and data might temporarily appear unavailable to your users.
It is recommended to shut down public access to the cluster while rebalancing
is in progress. In theory, no data should be lost as rebalancing is a
routine operation for Garage, although we cannot guarantee you that everything
will go right in such an extreme scenario.
#### `consistency_mode` {#consistency_mode}
The consistency mode setting determines the read and write behaviour of your cluster.
- `consistent`: The default setting. This is what the paragraph above describes.
The read and write quorum will be determined so that read-after-write consistency
is guaranteed.
- `degraded`: Lowers the read
quorum to `1`, to allow you to read data from your cluster when several
nodes (or nodes in several zones) are unavailable. In this mode, Garage
does not provide read-after-write consistency anymore.
The write quorum stays the same as in the `consistent` mode, ensuring that
data successfully written to Garage is stored on multiple nodes (depending
the replication factor).
- `dangerous`: This mode lowers both the read
and write quorums to `1`, to allow you to both read and write to your
cluster when several nodes (or nodes in several zones) are unavailable. It
is the least consistent mode of operation proposed by Garage, and also one
that should probably never be used.
Changing the `consistency_mode` between modes while leaving the `replication_factor` untouched
(e.g. setting your node's `consistency_mode` to `degraded` when it was previously unset, or from
`dangerous` to `consistent`), can be done easily by just changing the `consistency_mode`
parameter in your config files and restarting all your Garage nodes.
The consistency mode can be used together with various replication factors, to achieve
a wide range of read and write characteristics. Some examples:
- Replication factor `2`, consistency mode `degraded`: While this mode
technically exists, its properties are the same as with consistency mode `consistent`,
since the read quorum with replication factor `2`, consistency mode `consistent` is already 1.
- Replication factor `2`, consistency mode `dangerous`: written objects are written to
the second replica asynchronously. This means that Garage will return `200
OK` to a PutObject request before the second copy is fully written (or even
before it even starts being written). This means that data can more easily
be lost if the node crashes before a second copy can be completed. This
also means that written objects might not be visible immediately in read
operations. In other words, this configuration severely breaks the consistency and
durability guarantees of standard Garage cluster operation. Benefits of
this configuration: you can still write to your cluster when one node is
unavailable.
The quorums associated with each replication mode are described below:
| `replication_mode` | Number of replicas | Write quorum | Read quorum | Read-after-write consistency? |
| ------------------ | ------------------ | ------------ | ----------- | ----------------------------- |
| `none` or `1` | 1 | 1 | 1 | yes |
| `2` | 2 | 2 | 1 | yes |
| `2-dangerous` | 2 | 1 | 1 | NO |
| `3` | 3 | 2 | 2 | yes |
| `3-degraded` | 3 | 2 | 1 | NO |
| `3-dangerous` | 3 | 1 | 1 | NO |
| `consistency_mode` | `replication_factor` | Write quorum | Read quorum | Read-after-write consistency? |
| ------------------ | -------------------- | ------------ | ----------- | ----------------------------- |
| `consistent` | 1 | 1 | 1 | yes |
| `consistent` | 2 | 2 | 1 | yes |
| `dangerous` | 2 | 1 | 1 | NO |
| `consistent` | 3 | 2 | 2 | yes |
| `degraded` | 3 | 2 | 1 | NO |
| `dangerous` | 3 | 1 | 1 | NO |
Changing the `replication_mode` between modes with the same number of replicas
(e.g. from `3` to `3-degraded`, or from `2-dangerous` to `2`), can be done easily by
just changing the `replication_mode` parameter in your config files and restarting all your
Garage nodes.
#### `metadata_dir` {#metadata_dir}
It is also technically possible to change the replication mode to a mode with a
different numbers of replicas, although it's a dangerous operation that is not
officially supported. This requires you to delete the existing cluster layout
and create a new layout from scratch, meaning that a full rebalancing of your
cluster's data will be needed. To do it, shut down your cluster entirely,
delete the `custer_layout` files in the meta directories of all your nodes,
update all your configuration files with the new `replication_mode` parameter,
restart your cluster, and then create a new layout with all the nodes you want
to keep. Rebalancing data will take some time, and data might temporarily
appear unavailable to your users. It is recommended to shut down public access
to the cluster while rebalancing is in progress. In theory, no data should be
lost as rebalancing is a routine operation for Garage, although we cannot
guarantee you that everything will go right in such an extreme scenario.
The directory in which Garage will store its metadata. This contains the node identifier,
the network configuration and the peer list, the list of buckets and keys as well
as the index of all objects, object version and object blocks.
### `compression_level`
Store this folder on a fast SSD drive if possible to maximize Garage's performance.
#### `data_dir` {#data_dir}
The directory in which Garage will store the data blocks of objects.
This folder can be placed on an HDD. The space available for `data_dir`
should be counted to determine a node's capacity
when [adding it to the cluster layout](@/documentation/cookbook/real-world.md).
Since `v0.9.0`, Garage supports multiple data directories with the following syntax:
```toml
data_dir = [
{ path = "/path/to/old_data", read_only = true },
{ path = "/path/to/new_hdd1", capacity = "2T" },
{ path = "/path/to/new_hdd2", capacity = "4T" },
]
```
See [the dedicated documentation page](@/documentation/operations/multi-hdd.md)
on how to operate Garage in such a setup.
#### `db_engine` (since `v0.8.0`) {#db_engine}
Since `v0.8.0`, Garage can use alternative storage backends as follows:
| DB engine | `db_engine` value | Database path |
| --------- | ----------------- | ------------- |
| [LMDB](https://www.symas.com/lmdb) (since `v0.8.0`, default since `v0.9.0`) | `"lmdb"` | `<metadata_dir>/db.lmdb/` |
| [Sqlite](https://sqlite.org) (since `v0.8.0`) | `"sqlite"` | `<metadata_dir>/db.sqlite` |
| [Sled](https://sled.rs) (old default, removed since `v1.0`) | `"sled"` | `<metadata_dir>/db/` |
Sled was supported until Garage v0.9.x, and was removed in Garage v1.0.
You can still use an older binary of Garage (e.g. v0.9.4) to migrate
old Sled metadata databases to another engine.
Performance characteristics of the different DB engines are as follows:
- LMDB: the recommended database engine for high-performance distributed clusters.
LMDB works very well, but is known to have the following limitations:
- The data format of LMDB is not portable between architectures, so for
instance the Garage database of an x86-64 node cannot be moved to an ARM64
node.
- While LMDB can technically be used on 32-bit systems, this will limit your
node to very small database sizes due to how LMDB works; it is therefore
not recommended.
- Several users have reported corrupted LMDB database files after an unclean
shutdown (e.g. a power outage). This situation can generally be recovered
from if your cluster is geo-replicated (by rebuilding your metadata db from
other nodes), or if you have saved regular snapshots at the filesystem
level.
- Keys in LMDB are limited to 511 bytes. This limit translates to limits on
object keys in S3 and sort keys in K2V that are limted to 479 bytes.
- Sqlite: Garage supports Sqlite as an alternative storage backend for
metadata, which does not have the issues listed above for LMDB.
On versions 0.8.x and earlier, Sqlite should be avoided due to abysmal
performance, which was fixed with the addition of `metadata_fsync`.
Sqlite is still probably slower than LMDB due to the way we use it,
so it is not the best choice for high-performance storage clusters,
but it should work fine in many cases.
It is possible to convert Garage's metadata directory from one format to another
using the `garage convert-db` command, which should be used as follows:
```
garage convert-db -a <input db engine> -i <input db path> \
-b <output db engine> -o <output db path>
```
Make sure to specify the full database path as presented in the table above
(third colummn), and not just the path to the metadata directory.
#### `metadata_fsync` {#metadata_fsync}
Whether to enable synchronous mode for the database engine or not.
This is disabled (`false`) by default.
This reduces the risk of metadata corruption in case of power failures,
at the cost of a significant drop in write performance,
as Garage will have to pause to sync data to disk much more often
(several times for API calls such as PutObject).
Using this option reduces the risk of simultaneous metadata corruption on several
cluster nodes, which could lead to data loss.
If multi-site replication is used, this option is most likely not necessary, as
it is extremely unlikely that two nodes in different locations will have a
power failure at the exact same time.
(Metadata corruption on a single node is not an issue, the corrupted data file
can always be deleted and reconstructed from the other nodes in the cluster.)
Here is how this option impacts the different database engines:
| Database | `metadata_fsync = false` (default) | `metadata_fsync = true` |
|----------|------------------------------------|-------------------------------|
| Sqlite | `PRAGMA synchronous = OFF` | `PRAGMA synchronous = NORMAL` |
| LMDB | `MDB_NOMETASYNC` + `MDB_NOSYNC` | `MDB_NOMETASYNC` |
Note that the Sqlite database is always ran in `WAL` mode (`PRAGMA journal_mode = WAL`).
#### `data_fsync` {#data_fsync}
Whether to `fsync` data blocks and their containing directory after they are
saved to disk.
This is disabled (`false`) by default.
This might reduce the risk that a data block is lost in rare
situations such as simultaneous node losing power,
at the cost of a moderate drop in write performance.
Similarly to `metatada_fsync`, this is likely not necessary
if geographical replication is used.
#### `metadata_auto_snapshot_interval` (since Garage v0.9.4) {#metadata_auto_snapshot_interval}
If this value is set, Garage will automatically take a snapshot of the metadata
DB file at a regular interval and save it in the metadata directory.
This parameter can take any duration string that can be parsed by
the [`parse_duration`](https://docs.rs/parse_duration/latest/parse_duration/#syntax) crate.
Snapshots can allow to recover from situations where the metadata DB file is
corrupted, for instance after an unclean shutdown. See [this
page](@/documentation/operations/recovering.md#corrupted_meta) for details.
Garage keeps only the two most recent snapshots of the metadata DB and deletes
older ones automatically.
Note that taking a metadata snapshot is a relatively intensive operation as the
entire data file is copied. A snapshot being taken might have performance
impacts on the Garage node while it is running. If the cluster is under heavy
write load when a snapshot operation is running, this might also cause the
database file to grow in size significantly as pages cannot be recycled easily.
For this reason, it might be better to use filesystem-level snapshots instead
if possible.
#### `disable_scrub` {#disable_scrub}
By default, Garage runs a scrub of the data directory approximately once per
month, with a random delay to avoid all nodes running at the same time. When
it scrubs the data directory, Garage will read all of the data files stored on
disk to check their integrity, and will rebuild any data files that it finds
corrupted, using the remaining valid copies stored on other nodes.
See [this page](@/documentation/operations/durability-repairs.md#scrub) for details.
Set the `disable_scrub` configuration value to `true` if you don't need Garage
to scrub the data directory, for instance if you are already scrubbing at the
filesystem level. Note that in this case, if you find a corrupted data file,
you should delete it from the data directory and then call `garage repair
blocks` on the node to ensure that it re-obtains a copy from another node on
the network.
#### `use_local_tz` {#use_local_tz}
By default, Garage runs the lifecycle worker every day at midnight in UTC. Set the
`use_local_tz` configuration value to `true` if you want Garage to run the
lifecycle worker at midnight in your local timezone. If you have multiple nodes,
you should also ensure that each node has the same timezone configuration.
#### `block_size` {#block_size}
Garage splits stored objects in consecutive chunks of size `block_size`
(except the last one which might be smaller). The default size is 1MiB and
should work in most cases. We recommend increasing it to e.g. 10MiB if
you are using Garage to store large files and have fast network connections
between all nodes (e.g. 1gbps).
If you are interested in tuning this, feel free to do so (and remember to
report your findings to us!). When this value is changed for a running Garage
installation, only files newly uploaded will be affected. Previously uploaded
files will remain available. This however means that chunks from existing files
will not be deduplicated with chunks from newly uploaded files, meaning you
might use more storage space that is optimally possible.
#### `block_ram_buffer_max` (since v0.9.4) {#block_ram_buffer_max}
A limit on the total size of data blocks kept in RAM by S3 API nodes awaiting
to be sent to storage nodes asynchronously.
Explanation: since Garage wants to tolerate node failures, it uses quorum
writes to send data blocks to storage nodes: try to write the block to three
nodes, and return ok as soon as two writes complete. So even if all three nodes
are online, the third write always completes asynchronously. In general, there
are not many writes to a cluster, and the third asynchronous write can
terminate early enough so as to not cause unbounded RAM growth. However, if
the S3 API node is continuously receiving large quantities of data and the
third node is never able to catch up, many data blocks will be kept buffered in
RAM as they are awaiting transfer to the third node.
The `block_ram_buffer_max` sets a limit to the size of buffers that can be kept
in RAM in this process. When the limit is reached, backpressure is applied
back to the S3 client.
Note that this only counts buffers that have arrived to a certain stage of
processing (received from the client + encrypted and/or compressed as
necessary) and are ready to send to the storage nodes. Many other buffers will
not be counted and this is not a hard limit on RAM consumption. In particular,
if many clients send requests simultaneously with large objects, the RAM
consumption will always grow linearly with the number of concurrent requests,
as each request will use a few buffers of size `block_size` for receiving and
intermediate processing before even trying to send the data to the storage
node.
The default value is 256MiB.
#### `lmdb_map_size` {#lmdb_map_size}
This parameters can be used to set the map size used by LMDB,
which is the size of the virtual memory region used for mapping the database file.
The value of this parameter is the maximum size the metadata database can take.
This value is not bound by the physical RAM size of the machine running Garage.
If not specified, it defaults to 1GiB on 32-bit machines and 1TiB on 64-bit machines.
#### `compression_level` {#compression_level}
Zstd compression level to use for storing blocks.
@ -261,7 +514,7 @@ Compression is done synchronously, setting a value too high will add latency to
This value can be different between nodes, compression is done by the node which receive the
API call.
### `rpc_secret`, `rpc_secret_file` or `GARAGE_RPC_SECRET` (env)
#### `rpc_secret`, `rpc_secret_file` or `GARAGE_RPC_SECRET`, `GARAGE_RPC_SECRET_FILE` (env) {#rpc_secret}
Garage uses a secret key, called an RPC secret, that is shared between all
nodes of the cluster in order to identify these nodes and allow them to
@ -273,7 +526,10 @@ Since Garage `v0.8.2`, the RPC secret can also be stored in a file whose path is
given in the configuration variable `rpc_secret_file`, or specified as an
environment variable `GARAGE_RPC_SECRET`.
### `rpc_bind_addr`
Since Garage `v0.8.5` and `v0.9.1`, you can also specify the path of a file
storing the secret as the `GARAGE_RPC_SECRET_FILE` environment variable.
#### `rpc_bind_addr` {#rpc_bind_addr}
The address and port on which to bind for inter-cluster communcations
(reffered to as RPC for remote procedure calls).
@ -282,14 +538,33 @@ the node, even in the case of a NAT: the NAT should be configured to forward the
port number to the same internal port nubmer. This means that if you have several nodes running
behind a NAT, they should each use a different RPC port number.
### `rpc_public_addr`
#### `rpc_bind_outgoing`(since v0.9.2) {#rpc_bind_outgoing}
If enabled, pre-bind all sockets for outgoing connections to the same IP address
used for listening (the IP address specified in `rpc_bind_addr`) before
trying to connect to remote nodes.
This can be necessary if a node has multiple IP addresses,
but only one is allowed or able to reach the other nodes,
for instance due to firewall rules or specific routing configuration.
Disabled by default.
#### `rpc_public_addr` {#rpc_public_addr}
The address and port that other nodes need to use to contact this node for
RPC calls. **This parameter is optional but recommended.** In case you have
a NAT that binds the RPC port to a port that is different on your public IP,
this field might help making it work.
### `bootstrap_peers`
#### `rpc_public_addr_subnet` {#rpc_public_addr_subnet}
In case `rpc_public_addr` is not set, but autodiscovery is used, this allows
filtering the list of automatically discovered IPs to a specific subnet.
For example, if nodes should pick *their* IP inside a specific subnet, but you
don't want to explicitly write the IP down (as it's dynamic, or you want to
share configs across nodes), you can use this option.
#### `bootstrap_peers` {#bootstrap_peers}
A list of peer identifiers on which to contact other Garage peers of this cluster.
These peer identifiers have the following syntax:
@ -305,75 +580,118 @@ be obtained by running `garage node id` and then included directly in the
key will be returned by `garage node id` and you will have to add the IP
yourself.
### `allow_world_readable_secrets` or `GARAGE_ALLOW_WORLD_READABLE_SECRETS` (env) {#allow_world_readable_secrets}
## The `[consul_discovery]` section
Garage checks the permissions of your secret files to make sure they're not
world-readable. In some cases, the check might fail and consider your files as
world-readable even if they're not, for instance when using Posix ACLs.
Setting `allow_world_readable_secrets` to `true` bypass this
permission verification.
Alternatively, you can set the `GARAGE_ALLOW_WORLD_READABLE_SECRETS`
environment variable to `true` to bypass the permissions check.
### The `[consul_discovery]` section
Garage supports discovering other nodes of the cluster using Consul. For this
to work correctly, nodes need to know their IP address by which they can be
reached by other nodes of the cluster, which should be set in `rpc_public_addr`.
### `consul_http_addr` and `service_name`
#### `consul_http_addr` {#consul_http_addr}
The `consul_http_addr` parameter should be set to the full HTTP(S) address of the Consul server.
### `service_name`
#### `api` {#consul_api}
Two APIs for service registration are supported: `catalog` and `agent`. `catalog`, the default, will register a service using
the `/v1/catalog` endpoints, enabling mTLS if `client_cert` and `client_key` are provided. The `agent` API uses the
`v1/agent` endpoints instead, where an optional `token` may be provided.
#### `service_name` {#consul_service_name}
`service_name` should be set to the service name under which Garage's
RPC ports are announced.
### `client_cert`, `client_key`
#### `client_cert`, `client_key` {#consul_client_cert_and_key}
TLS client certificate and client key to use when communicating with Consul over TLS. Both are mandatory when doing so.
Only available when `api = "catalog"`.
### `ca_cert`
#### `ca_cert` {#consul_ca_cert}
TLS CA certificate to use when communicating with Consul over TLS.
### `tls_skip_verify`
#### `tls_skip_verify` {#consul_tls_skip_verify}
Skip server hostname verification in TLS handshake.
`ca_cert` is ignored when this is set.
#### `token` {#consul_token}
## The `[kubernetes_discovery]` section
Uses the provided token for communication with Consul. Only available when `api = "agent"`.
The policy assigned to this token should at least have these rules:
```hcl
// the `service_name` specified above
service "garage" {
policy = "write"
}
service_prefix "" {
policy = "read"
}
node_prefix "" {
policy = "read"
}
```
#### `tags` and `meta` {#consul_tags_and_meta}
Additional list of tags and map of service meta to add during service registration.
### The `[kubernetes_discovery]` section
Garage supports discovering other nodes of the cluster using kubernetes custom
resources. For this to work, a `[kubernetes_discovery]` section must be present
with at least the `namespace` and `service_name` parameters.
### `namespace`
#### `namespace` {#kube_namespace}
`namespace` sets the namespace in which the custom resources are
configured.
### `service_name`
#### `service_name` {#kube_service_name}
`service_name` is added as a label to the advertised resources to
filter them, to allow for multiple deployments in a single namespace.
### `skip_crd`
#### `skip_crd` {#kube_skip_crd}
`skip_crd` can be set to true to disable the automatic creation and
patching of the `garagenodes.deuxfleurs.fr` CRD. You will need to create the CRD
manually.
## The `[s3_api]` section
### The `[s3_api]` section
### `api_bind_addr`
#### `api_bind_addr` {#s3_api_bind_addr}
The IP and port on which to bind for accepting S3 API calls.
This endpoint does not suport TLS: a reverse proxy should be used to provide it.
### `s3_region`
Alternatively, since `v0.8.5`, a path can be used to create a unix socket with 0222 mode.
#### `s3_region` {#s3_region}
Garage will accept S3 API calls that are targetted to the S3 region defined here.
API calls targetted to other regions will fail with a AuthorizationHeaderMalformed error
message that redirects the client to the correct region.
### `root_domain` {#root_domain}
#### `root_domain` {#s3_root_domain}
The optionnal suffix to access bucket using vhost-style in addition to path-style request.
The optional suffix to access bucket using vhost-style in addition to path-style request.
Note path-style requests are always enabled, whether or not vhost-style is configured.
Configuring vhost-style S3 required a wildcard DNS entry, and possibly a wildcard TLS certificate,
but might be required by softwares not supporting path-style requests.
@ -383,59 +701,66 @@ using the hostname `my-bucket.s3.garage.eu`.
## The `[s3_web]` section
### The `[s3_web]` section
Garage allows to publish content of buckets as websites. This section configures the
behaviour of this module.
### `bind_addr`
#### `bind_addr` {#web_bind_addr}
The IP and port on which to bind for accepting HTTP requests to buckets configured
for website access.
This endpoint does not suport TLS: a reverse proxy should be used to provide it.
### `root_domain`
Alternatively, since `v0.8.5`, a path can be used to create a unix socket with 0222 mode.
The optionnal suffix appended to bucket names for the corresponding HTTP Host.
#### `root_domain` {#web_root_domain}
The optional suffix appended to bucket names for the corresponding HTTP Host.
For instance, if `root_domain` is `web.garage.eu`, a bucket called `deuxfleurs.fr`
will be accessible either with hostname `deuxfleurs.fr.web.garage.eu`
or with hostname `deuxfleurs.fr`.
## The `[admin]` section
### The `[admin]` section
Garage has a few administration capabilities, in particular to allow remote monitoring. These features are detailed below.
### `api_bind_addr`
#### `api_bind_addr` {#admin_api_bind_addr}
If specified, Garage will bind an HTTP server to this port and address, on
which it will listen to requests for administration features.
See [administration API reference](@/documentation/reference-manual/admin-api.md) to learn more about these features.
### `metrics_token`, `metrics_token_file` or `GARAGE_METRICS_TOKEN` (env)
Alternatively, since `v0.8.5`, a path can be used to create a unix socket. Note that for security reasons,
the socket will have 0220 mode. Make sure to set user and group permissions accordingly.
#### `metrics_token`, `metrics_token_file` or `GARAGE_METRICS_TOKEN`, `GARAGE_METRICS_TOKEN_FILE` (env) {#admin_metrics_token}
The token for accessing the Metrics endpoint. If this token is not set, the
Metrics endpoint can be accessed without access control.
You can use any random string for this value. We recommend generating a random token with `openssl rand -hex 32`.
You can use any random string for this value. We recommend generating a random token with `openssl rand -base64 32`.
`metrics_token` was introduced in Garage `v0.7.2`.
`metrics_token_file` and the `GARAGE_METRICS_TOKEN` environment variable are supported since Garage `v0.8.2`.
`GARAGE_METRICS_TOKEN_FILE` is supported since `v0.8.5` / `v0.9.1`.
### `admin_token`, `admin_token_file` or `GARAGE_ADMIN_TOKEN` (env)
#### `admin_token`, `admin_token_file` or `GARAGE_ADMIN_TOKEN`, `GARAGE_ADMIN_TOKEN_FILE` (env) {#admin_token}
The token for accessing all of the other administration endpoints. If this
token is not set, access to these endpoints is disabled entirely.
You can use any random string for this value. We recommend generating a random token with `openssl rand -hex 32`.
You can use any random string for this value. We recommend generating a random token with `openssl rand -base64 32`.
`admin_token` was introduced in Garage `v0.7.2`.
`admin_token_file` and the `GARAGE_ADMIN_TOKEN` environment variable are supported since Garage `v0.8.2`.
`GARAGE_ADMIN_TOKEN_FILE` is supported since `v0.8.5` / `v0.9.1`.
### `trace_sink`
#### `trace_sink` {#admin_trace_sink}
Optionally, the address of an OpenTelemetry collector. If specified,
Garage will send traces in the OpenTelemetry format to this endpoint. These

View file

@ -35,7 +35,22 @@ This makes setting up and administering storage clusters, we hope, as easy as it
A Garage cluster can very easily evolve over time, as storage nodes are added or removed.
Garage will automatically rebalance data between nodes as needed to ensure the desired number of copies.
Read about cluster layout management [here](@/documentation/reference-manual/layout.md).
Read about cluster layout management [here](@/documentation/operations/layout.md).
### Several replication modes
Garage supports a variety of replication modes, with configurable replica count,
and with various levels of consistency, in order to adapt to a variety of usage scenarios.
Read our reference page on [supported replication modes](@/documentation/reference-manual/configuration.md#replication_factor)
to select the replication mode best suited to your use case (hint: in most cases, `replication_factor = 3` is what you want).
### Compression and deduplication
All data stored in Garage is deduplicated, and optionnally compressed using
Zstd. Objects uploaded to Garage are chunked in blocks of constant sizes (see
[`block_size`](@/documentation/reference-manual/configuration.md#block_size)),
and the hashes of individual blocks are used to dispatch them to storage nodes
and to deduplicate them.
### No RAFT slowing you down
@ -48,13 +63,6 @@ As a consequence, requests can be handled much faster, even in cases where laten
between cluster nodes is important (see our [benchmarks](@/documentation/design/benchmarks/index.md) for data on this).
This is particularly usefull when nodes are far from one another and talk to one other through standard Internet connections.
### Several replication modes
Garage supports a variety of replication modes, with 1 copy, 2 copies or 3 copies of your data,
and with various levels of consistency, in order to adapt to a variety of usage scenarios.
Read our reference page on [supported replication modes](@/documentation/reference-manual/configuration.md#replication-mode)
to select the replication mode best suited to your use case (hint: in most cases, `replication_mode = "3"` is what you want).
### Web server for static websites
A storage bucket can easily be configured to be served directly by Garage as a static web site.

View file

@ -3,7 +3,7 @@ title = "K2V"
weight = 100
+++
Starting with version 0.7.2, Garage introduces an optionnal feature, K2V,
Starting with version 0.7.2, Garage introduces an optional feature, K2V,
which is an alternative storage API designed to help efficiently store
many small values in buckets (in opposition to S3 which is more designed
to store large blobs).

View file

@ -1,77 +0,0 @@
+++
title = "Cluster layout management"
weight = 50
+++
The cluster layout in Garage is a table that assigns to each node a role in
the cluster. The role of a node in Garage can either be a storage node with
a certain capacity, or a gateway node that does not store data and is only
used as an API entry point for faster cluster access.
An introduction to building cluster layouts can be found in the [production deployment](@/documentation/cookbook/real-world.md) page.
## How cluster layouts work in Garage
In Garage, a cluster layout is composed of the following components:
- a table of roles assigned to nodes
- a version number
Garage nodes will always use the cluster layout with the highest version number.
Garage nodes also maintain and synchronize between them a set of proposed role
changes that haven't yet been applied. These changes will be applied (or
canceled) in the next version of the layout
The following commands insert modifications to the set of proposed role changes
for the next layout version (but they do not create the new layout immediately):
```bash
garage layout assign [...]
garage layout remove [...]
```
The following command can be used to inspect the layout that is currently set in the cluster
and the changes proposed for the next layout version, if any:
```bash
garage layout show
```
The following commands create a new layout with the specified version number,
that either takes into account the proposed changes or cancels them:
```bash
garage layout apply --version <new_version_number>
garage layout revert --version <new_version_number>
```
The version number of the new layout to create must be 1 + the version number
of the previous layout that existed in the cluster. The `apply` and `revert`
commands will fail otherwise.
## Warnings about Garage cluster layout management
**Warning: never make several calls to `garage layout apply` or `garage layout
revert` with the same value of the `--version` flag. Doing so can lead to the
creation of several different layouts with the same version number, in which
case your Garage cluster will become inconsistent until fixed.** If a call to
`garage layout apply` or `garage layout revert` has failed and `garage layout
show` indicates that a new layout with the given version number has not been
set in the cluster, then it is fine to call the command again with the same
version number.
If you are using the `garage` CLI by typing individual commands in your
shell, you shouldn't have much issues as long as you run commands one after
the other and take care of checking the output of `garage layout show`
before applying any changes.
If you are using the `garage` CLI to script layout changes, follow the following recommendations:
- Make all of your `garage` CLI calls to the same RPC host. Do not use the
`garage` CLI to connect to individual nodes to send them each a piece of the
layout changes you are making, as the changes propagate asynchronously
between nodes and might not all be taken into account at the time when the
new layout is applied.
- **Only call `garage layout apply` once**, and call it **strictly after** all
of the `layout assign` and `layout remove` commands have returned.

View file

@ -27,6 +27,112 @@ Exposes the Garage replication factor configured on the node
garage_replication_factor 3
```
#### `garage_local_disk_avail` and `garage_local_disk_total` (gauge)
Reports the available and total disk space on each node, for data and metadata separately.
```
garage_local_disk_avail{volume="data"} 540341960704
garage_local_disk_avail{volume="metadata"} 540341960704
garage_local_disk_total{volume="data"} 763063566336
garage_local_disk_total{volume="metadata"} 763063566336
```
### Cluster health status metrics
#### `cluster_healthy` (gauge)
Whether all storage nodes are connected (0 or 1)
```
cluster_healthy 0
```
#### `cluster_available` (gauge)
Whether all requests can be served, even if some storage nodes are disconnected
```
cluster_available 1
```
#### `cluster_connected_nodes` (gauge)
Number of nodes currently connected
```
cluster_connected_nodes 3
```
#### `cluster_known_nodes` (gauge)
Number of nodes already seen once in the cluster
```
cluster_known_nodes 3
```
#### `cluster_layout_node_connected` (gauge)
Connection status for individual nodes of the cluster layout
```
cluster_layout_node_connected{id="62b218d848e86a64",role_capacity="1000000000",role_gateway="0",role_zone="dc1"} 1
cluster_layout_node_connected{id="a11c7cf18af29737",role_capacity="1000000000",role_gateway="0",role_zone="dc1"} 0
cluster_layout_node_connected{id="a235ac7695e0c54d",role_capacity="1000000000",role_gateway="0",role_zone="dc1"} 1
cluster_layout_node_connected{id="b10c110e4e854e5a",role_capacity="1000000000",role_gateway="0",role_zone="dc1"} 1
```
#### `cluster_layout_node_disconnected_time` (gauge)
Time (in seconds) since last connection to individual nodes of the cluster layout
```
cluster_layout_node_disconnected_time{id="62b218d848e86a64",role_capacity="1000000000",role_gateway="0",role_zone="dc1"} 0
cluster_layout_node_disconnected_time{id="a235ac7695e0c54d",role_capacity="1000000000",role_gateway="0",role_zone="dc1"} 0
cluster_layout_node_disconnected_time{id="b10c110e4e854e5a",role_capacity="1000000000",role_gateway="0",role_zone="dc1"} 0
```
#### `cluster_storage_nodes` (gauge)
Number of storage nodes declared in the current layout
```
cluster_storage_nodes 4
```
#### `cluster_storage_nodes_ok` (gauge)
Number of storage nodes currently connected
```
cluster_storage_nodes_ok 3
```
#### `cluster_partitions` (gauge)
Number of partitions in the layout (this is always 256)
```
cluster_partitions 256
```
#### `cluster_partitions_all_ok` (gauge)
Number of partitions for which all storage nodes are connected
```
cluster_partitions_all_ok 64
```
#### `cluster_partitions_quorum` (gauge)
Number of partitions for which we have a quorum of connected nodes and all requests can be served
```
cluster_partitions_quorum 256
```
### Metrics of the API endpoints
#### `api_admin_request_counter` (counter)
@ -119,6 +225,17 @@ block_bytes_read 120586322022
block_bytes_written 3386618077
```
#### `block_ram_buffer_free_kb` (gauge)
Kibibytes available for buffering blocks that have to be sent to remote nodes.
When clients send too much data to this node and a storage node is not receiving
data fast enough due to slower network conditions, this will decrease down to
zero and backpressure will be applied.
```
block_ram_buffer_free_kb 219829
```
#### `block_compression_level` (counter)
Exposes the block compression level configured for the Garage node.

View file

@ -33,6 +33,7 @@ Feel free to open a PR to suggest fixes this table. Minio is missing because the
| [URL path-style](https://docs.aws.amazon.com/AmazonS3/latest/userguide/VirtualHosting.html#path-style-access) (eg. `host.tld/bucket/key`) | ✅ Implemented | ✅ | ✅ | ❓| ✅ |
| [URL vhost-style](https://docs.aws.amazon.com/AmazonS3/latest/userguide/VirtualHosting.html#virtual-hosted-style-access) URL (eg. `bucket.host.tld/key`) | ✅ Implemented | ❌| ✅| ✅ | ✅ |
| [Presigned URLs](https://docs.aws.amazon.com/AmazonS3/latest/userguide/ShareObjectPreSignedURL.html) | ✅ Implemented | ❌| ✅ | ✅ | ✅(❓) |
| [SSE-C encryption](https://docs.aws.amazon.com/AmazonS3/latest/userguide/ServerSideEncryptionCustomerKeys.html) | ✅ Implemented | ❓ | ✅ | ❌ | ✅ |
*Note:* OpenIO does not says if it supports presigned URLs. Because it is part
of signature v4 and they claim they support it without additional precisions,
@ -76,16 +77,13 @@ but these endpoints are documented in [Red Hat Ceph Storage - Chapter 2. Ceph Ob
| Endpoint | Garage | [Openstack Swift](https://docs.openstack.org/swift/latest/s3_compat.html) | [Ceph Object Gateway](https://docs.ceph.com/en/latest/radosgw/s3/) | [Riak CS](https://docs.riak.com/riak/cs/2.1.1/references/apis/storage/s3/index.html) | [OpenIO](https://docs.openio.io/latest/source/arch-design/s3_compliancy.html) |
|------------------------------|----------------------------------|-----------------|---------------|---------|-----|
| [AbortMultipartUpload](https://docs.aws.amazon.com/AmazonS3/latest/API/API_AbortMultipartUpload.html) | ✅ Implemented | ✅ | ✅ | ✅ | ✅ |
| [CompleteMultipartUpload](https://docs.aws.amazon.com/AmazonS3/latest/API/API_CompleteMultipartUpload.html) | ✅ Implemented (see details below) | ✅ | ✅ | ✅ | ✅ |
| [CompleteMultipartUpload](https://docs.aws.amazon.com/AmazonS3/latest/API/API_CompleteMultipartUpload.html) | ✅ Implemented | ✅ | ✅ | ✅ | ✅ |
| [CreateMultipartUpload](https://docs.aws.amazon.com/AmazonS3/latest/API/API_CreateMultipartUpload.html) | ✅ Implemented | ✅| ✅ | ✅ | ✅ |
| [ListMultipartUpload](https://docs.aws.amazon.com/AmazonS3/latest/API/API_ListMultipartUpload.html) | ✅ Implemented | ✅ | ✅ | ✅ | ✅ |
| [ListParts](https://docs.aws.amazon.com/AmazonS3/latest/API/API_ListParts.html) | ✅ Implemented | ✅ | ✅ | ✅ | ✅ |
| [UploadPart](https://docs.aws.amazon.com/AmazonS3/latest/API/API_UploadPart.html) | ✅ Implemented (see details below) | ✅ | ✅| ✅ | ✅ |
| [UploadPart](https://docs.aws.amazon.com/AmazonS3/latest/API/API_UploadPart.html) | ✅ Implemented | ✅ | ✅| ✅ | ✅ |
| [UploadPartCopy](https://docs.aws.amazon.com/AmazonS3/latest/API/API_UploadPartCopy.html) | ✅ Implemented | ✅ | ✅ | ✅ | ✅ |
Our implementation of Multipart Upload is currently a bit more restrictive than Amazon's one in some edge cases.
For more information, please refer to our [issue tracker](https://git.deuxfleurs.fr/Deuxfleurs/garage/issues/204).
### Website endpoints
| Endpoint | Garage | [Openstack Swift](https://docs.openstack.org/swift/latest/s3_compat.html) | [Ceph Object Gateway](https://docs.ceph.com/en/latest/radosgw/s3/) | [Riak CS](https://docs.riak.com/riak/cs/2.1.1/references/apis/storage/s3/index.html) | [OpenIO](https://docs.openio.io/latest/source/arch-design/s3_compliancy.html) |
@ -127,15 +125,22 @@ If you need this feature, please [share your use case in our dedicated issue](ht
| Endpoint | Garage | [Openstack Swift](https://docs.openstack.org/swift/latest/s3_compat.html) | [Ceph Object Gateway](https://docs.ceph.com/en/latest/radosgw/s3/) | [Riak CS](https://docs.riak.com/riak/cs/2.1.1/references/apis/storage/s3/index.html) | [OpenIO](https://docs.openio.io/latest/source/arch-design/s3_compliancy.html) |
|------------------------------|----------------------------------|-----------------|---------------|---------|-----|
| [DeleteBucketLifecycle](https://docs.aws.amazon.com/AmazonS3/latest/API/API_DeleteBucketLifecycle.html) | ❌ Missing | ❌| ✅| ❌| ✅|
| [GetBucketLifecycleConfiguration](https://docs.aws.amazon.com/AmazonS3/latest/API/API_GetBucketLifecycleConfiguration.html) | ❌ Missing | ❌| ✅ | ❌| ✅|
| [PutBucketLifecycleConfiguration](https://docs.aws.amazon.com/AmazonS3/latest/API/API_PutBucketLifecycleConfiguration.html) | ❌ Missing | ❌| ✅ | ❌| ✅|
| [DeleteBucketLifecycle](https://docs.aws.amazon.com/AmazonS3/latest/API/API_DeleteBucketLifecycle.html) | ✅ Implemented | ❌| ✅| ❌| ✅|
| [GetBucketLifecycleConfiguration](https://docs.aws.amazon.com/AmazonS3/latest/API/API_GetBucketLifecycleConfiguration.html) | ✅ Implemented | ❌| ✅ | ❌| ✅|
| [PutBucketLifecycleConfiguration](https://docs.aws.amazon.com/AmazonS3/latest/API/API_PutBucketLifecycleConfiguration.html) | ⚠ Partially implemented (see below) | ❌| ✅ | ❌| ✅|
| [GetBucketVersioning](https://docs.aws.amazon.com/AmazonS3/latest/API/API_GetBucketVersioning.html) | ❌ Stub (see below) | ✅| ✅ | ❌| ✅|
| [ListObjectVersions](https://docs.aws.amazon.com/AmazonS3/latest/API/API_ListObjectVersions.html) | ❌ Missing | ❌| ✅ | ❌| ✅|
| [PutBucketVersioning](https://docs.aws.amazon.com/AmazonS3/latest/API/API_PutBucketVersioning.html) | ❌ Missing | ❌| ✅| ❌| ✅|
**PutBucketLifecycleConfiguration:** The only actions supported are
`AbortIncompleteMultipartUpload` and `Expiration` (without the
`ExpiredObjectDeleteMarker` field). All other operations are dependent on
either bucket versionning or storage classes which Garage currently does not
implement. The deprecated `Prefix` member directly in the the `Rule`
structure/XML tag is not supported, specified prefixes must be inside the
`Filter` structure/XML tag.
**GetBucketVersioning:** Stub implementation (Garage does not yet support versionning so this always returns "versionning not enabled").
**GetBucketVersioning:** Stub implementation which always returns "versionning not enabled", since Garage does not yet support bucket versionning.
### Replication endpoints

View file

@ -1,6 +1,6 @@
+++
title = "Working Documents"
weight = 8
weight = 90
sort_by = "weight"
template = "documentation.html"
+++

View file

@ -0,0 +1,72 @@
+++
title = "Migrating from 0.8 to 0.9"
weight = 12
+++
**This guide explains how to migrate to 0.9 if you have an existing 0.8 cluster.
We don't recommend trying to migrate to 0.9 directly from 0.7 or older.**
This migration procedure has been tested on several clusters without issues.
However, it is still a *critical procedure* that might cause issues.
**Make sure to back up all your data before attempting it!**
You might also want to read our [general documentation on upgrading Garage](@/documentation/operations/upgrading.md).
The following are **breaking changes** in Garage v0.9 that require your attention when migrating:
- LMDB is now the default metadata db engine and Sled is deprecated. If you were using Sled, make sure to specify `db_engine = "sled"` in your configuration file, or take the time to [convert your database](https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#db-engine-since-v0-8-0).
- Capacity values are now in actual byte units. The translation from the old layout will assign 1 capacity = 1Gb by default, which might be wrong for your cluster. This does not cause any data to be moved around, but you might want to re-assign correct capacity values post-migration.
- Multipart uploads that were started in Garage v0.8 will not be visible in Garage v0.9 and will have to be restarted from scratch.
- Changes to the admin API: some `v0/` endpoints have been replaced by `v1/` counterparts with updated/uniformized syntax. All other endpoints have also moved to `v1/` by default, without syntax changes, but are still available under `v0/` for compatibility.
## Simple migration procedure (takes cluster offline for a while)
The migration steps are as follows:
1. Disable API and web access. You may do this by stopping your reverse proxy or by commenting out
the `api_bind_addr` values in your `config.toml` file and restarting Garage.
2. Do `garage repair --all-nodes --yes tables` and `garage repair --all-nodes --yes blocks`,
check the logs and check that all data seems to be synced correctly between
nodes. If you have time, do additional checks (`versions`, `block_refs`, etc.)
3. Check that the block resync queue and Merkle queue are empty:
run `garage stats -a` to query them or inspect metrics in the Grafana dashboard.
4. Turn off Garage v0.8
5. **Backup the metadata folder of all your nodes!** For instance, use the following command
if your metadata directory is `/var/lib/garage/meta`: `cd /var/lib/garage ; tar -acf meta-v0.8.tar.zst meta/`
6. Install Garage v0.9
7. Update your configuration file if necessary.
8. Turn on Garage v0.9
9. Do `garage repair --all-nodes --yes tables` and `garage repair --all-nodes --yes blocks`.
Wait for a full table sync to run.
10. Your upgraded cluster should be in a working state. Re-enable API and Web
access and check that everything went well.
11. Monitor your cluster in the next hours to see if it works well under your production load, report any issue.
12. You might want to assign correct capacity values to all your nodes. Doing so might cause data to be moved
in your cluster, which should also be monitored carefully.
## Minimal downtime migration procedure
The migration to Garage v0.9 can be done with almost no downtime,
by restarting all nodes at once in the new version.
The migration steps are as follows:
1. Do `garage repair --all-nodes --yes tables` and `garage repair --all-nodes --yes blocks`,
check the logs and check that all data seems to be synced correctly between
nodes. If you have time, do additional checks (`versions`, `block_refs`, etc.)
2. Turn off each node individually; back up its metadata folder (see above); turn it back on again.
This will allow you to take a backup of all nodes without impacting global cluster availability.
You can do all nodes of a single zone at once as this does not impact the availability of Garage.
3. Prepare your binaries and configuration files for Garage v0.9
4. Shut down all v0.8 nodes simultaneously, and restart them all simultaneously in v0.9.
Use your favorite deployment tool (Ansible, Kubernetes, Nomad) to achieve this as fast as possible.
Garage v0.9 should be in a working state as soon as it starts.
5. Proceed with repair and monitoring as described in steps 9-12 above.

View file

@ -0,0 +1,77 @@
+++
title = "Migrating from 0.9 to 1.0"
weight = 11
+++
**This guide explains how to migrate to 1.0 if you have an existing 0.9 cluster.
We don't recommend trying to migrate to 1.0 directly from 0.8 or older.**
This migration procedure has been tested on several clusters without issues.
However, it is still a *critical procedure* that might cause issues.
**Make sure to back up all your data before attempting it!**
You might also want to read our [general documentation on upgrading Garage](@/documentation/operations/upgrading.md).
## Changes introduced in v1.0
The following are **breaking changes** in Garage v1.0 that require your attention when migrating:
- The Sled metadata db engine has been **removed**. If your cluster was still
using Sled, you will need to **use a Garage v0.9.x binary** to convert the
database using the `garage convert-db` subcommand. See
[here](@/documentation/reference-manual/configuration.md#db_engine) for the
details of the procedure.
The following syntax changes have been made to the configuration file:
- The `replication_mode` parameter has been split into two parameters:
[`replication_factor`](@/documentation/reference-manual/configuration.md#replication_factor)
and
[`consistency_mode`](@/documentation/reference-manual/configuration.md#consistency_mode).
The old syntax using `replication_mode` is still supported for legacy
reasons and can still be used.
- The parameters `sled_cache_capacity` and `sled_flush_every_ms` have been removed.
## Migration procedure
The migration to Garage v1.0 can be done with almost no downtime,
by restarting all nodes at once in the new version.
The migration steps are as follows:
1. Do a `garage repair --all-nodes --yes tables`, check the logs and check that
all data seems to be synced correctly between nodes. If you have time, do
additional `garage repair` procedures (`blocks`, `versions`, `block_refs`,
etc.)
2. Ensure you have a snapshot of your Garage installation that you can restore
to in case the upgrade goes wrong:
- If you are running Garage v0.9.4 or later, use the `garage meta snapshot
--all` to make a backup snapshot of the metadata directories of your nodes
for backup purposes, and save a copy of the following files in the
metadata directories of your nodes: `cluster_layout`, `data_layout`,
`node_key`, `node_key.pub`.
- If you are running a filesystem such as ZFS or BTRFS that support
snapshotting, you can create a filesystem-level snapshot to be used as a
restoration point if needed.
- In other cases, make a backup using the old procedure: turn off each node
individually; back up its metadata folder (for instance, use the following
command if your metadata directory is `/var/lib/garage/meta`: `cd
/var/lib/garage ; tar -acf meta-v0.9.tar.zst meta/`); turn it back on
again. This will allow you to take a backup of all nodes without
impacting global cluster availability. You can do all nodes of a single
zone at once as this does not impact the availability of Garage.
3. Prepare your updated binaries and configuration files for Garage v1.0
4. Shut down all v0.9 nodes simultaneously, and restart them all simultaneously
in v1.0. Use your favorite deployment tool (Ansible, Kubernetes, Nomad) to
achieve this as fast as possible. Garage v1.0 should be in a working state
as soon as enough nodes have started.
5. Monitor your cluster in the following hours to see if it works well under
your production load.

View file

@ -8,9 +8,9 @@ listen address is specified in the `[admin]` section of the configuration
file (see [configuration file
reference](@/documentation/reference-manual/configuration.md))
**WARNING.** At this point, there is no comittement to stability of the APIs described in this document.
We will bump the version numbers prefixed to each API endpoint at each time the syntax
or semantics change, meaning that code that relies on these endpoint will break
**WARNING.** At this point, there is no commitment to the stability of the APIs described in this document.
We will bump the version numbers prefixed to each API endpoint each time the syntax
or semantics change, meaning that code that relies on these endpoints will break
when changes are introduced.
The Garage administration API was introduced in version 0.7.2, this document
@ -19,7 +19,7 @@ does not apply to older versions of Garage.
## Access control
The admin API uses two different tokens for acces control, that are specified in the config file's `[admin]` section:
The admin API uses two different tokens for access control, that are specified in the config file's `[admin]` section:
- `metrics_token`: the token for accessing the Metrics endpoint (if this token
is not set in the config file, the Metrics endpoint can be accessed without
@ -52,11 +52,11 @@ Returns an HTTP status 200 if the node is ready to answer user's requests,
and an HTTP status 503 (Service Unavailable) if there are some partitions
for which a quorum of nodes is not available.
A simple textual message is also returned in a body with content-type `text/plain`.
See `/v0/health` for an API that also returns JSON output.
See `/v1/health` for an API that also returns JSON output.
### Cluster operations
#### GetClusterStatus `GET /v0/status`
#### GetClusterStatus `GET /v1/status`
Returns the cluster's current status in JSON, including:
@ -69,87 +69,121 @@ Example response body:
```json
{
"node": "ec79480e0ce52ae26fd00c9da684e4fa56658d9c64cdcecb094e936de0bfe71f",
"garage_version": "git:v0.8.0",
"knownNodes": {
"ec79480e0ce52ae26fd00c9da684e4fa56658d9c64cdcecb094e936de0bfe71f": {
"addr": "10.0.0.11:3901",
"is_up": true,
"last_seen_secs_ago": 9,
"hostname": "node1"
},
"4a6ae5a1d0d33bf895f5bb4f0a418b7dc94c47c0dd2eb108d1158f3c8f60b0ff": {
"addr": "10.0.0.12:3901",
"is_up": true,
"last_seen_secs_ago": 1,
"hostname": "node2"
},
"23ffd0cdd375ebff573b20cc5cef38996b51c1a7d6dbcf2c6e619876e507cf27": {
"addr": "10.0.0.21:3901",
"is_up": true,
"last_seen_secs_ago": 7,
"hostname": "node3"
},
"e2ee7984ee65b260682086ec70026165903c86e601a4a5a501c1900afe28d84b": {
"addr": "10.0.0.22:3901",
"is_up": true,
"last_seen_secs_ago": 1,
"hostname": "node4"
}
},
"layout": {
"version": 12,
"roles": {
"ec79480e0ce52ae26fd00c9da684e4fa56658d9c64cdcecb094e936de0bfe71f": {
"node": "b10c110e4e854e5aa3f4637681befac755154b20059ec163254ddbfae86b09df",
"garageVersion": "v1.0.1",
"garageFeatures": [
"k2v",
"lmdb",
"sqlite",
"metrics",
"bundled-libs"
],
"rustVersion": "1.68.0",
"dbEngine": "LMDB (using Heed crate)",
"layoutVersion": 5,
"nodes": [
{
"id": "62b218d848e86a64f7fe1909735f29a4350547b54c4b204f91246a14eb0a1a8c",
"role": {
"id": "62b218d848e86a64f7fe1909735f29a4350547b54c4b204f91246a14eb0a1a8c",
"zone": "dc1",
"capacity": 4,
"tags": [
"node1"
]
"capacity": 100000000000,
"tags": []
},
"4a6ae5a1d0d33bf895f5bb4f0a418b7dc94c47c0dd2eb108d1158f3c8f60b0ff": {
"addr": "10.0.0.3:3901",
"hostname": "node3",
"isUp": true,
"lastSeenSecsAgo": 12,
"draining": false,
"dataPartition": {
"available": 660270088192,
"total": 873862266880
},
"metadataPartition": {
"available": 660270088192,
"total": 873862266880
}
},
{
"id": "a11c7cf18af297379eff8688360155fe68d9061654449ba0ce239252f5a7487f",
"role": null,
"addr": "10.0.0.2:3901",
"hostname": "node2",
"isUp": true,
"lastSeenSecsAgo": 11,
"draining": true,
"dataPartition": {
"available": 660270088192,
"total": 873862266880
},
"metadataPartition": {
"available": 660270088192,
"total": 873862266880
}
},
{
"id": "a235ac7695e0c54d7b403943025f57504d500fdcc5c3e42c71c5212faca040a2",
"role": {
"id": "a235ac7695e0c54d7b403943025f57504d500fdcc5c3e42c71c5212faca040a2",
"zone": "dc1",
"capacity": 6,
"tags": [
"node2"
]
"capacity": 100000000000,
"tags": []
},
"23ffd0cdd375ebff573b20cc5cef38996b51c1a7d6dbcf2c6e619876e507cf27": {
"zone": "dc2",
"capacity": 10,
"tags": [
"node3"
]
"addr": "127.0.0.1:3904",
"hostname": "lindy",
"isUp": true,
"lastSeenSecsAgo": 2,
"draining": false,
"dataPartition": {
"available": 660270088192,
"total": 873862266880
},
"metadataPartition": {
"available": 660270088192,
"total": 873862266880
}
},
"stagedRoleChanges": {
"e2ee7984ee65b260682086ec70026165903c86e601a4a5a501c1900afe28d84b": {
"zone": "dc2",
"capacity": 5,
"tags": [
"node4"
{
"id": "b10c110e4e854e5aa3f4637681befac755154b20059ec163254ddbfae86b09df",
"role": {
"id": "b10c110e4e854e5aa3f4637681befac755154b20059ec163254ddbfae86b09df",
"zone": "dc1",
"capacity": 100000000000,
"tags": []
},
"addr": "10.0.0.1:3901",
"hostname": "node1",
"isUp": true,
"lastSeenSecsAgo": 3,
"draining": false,
"dataPartition": {
"available": 660270088192,
"total": 873862266880
},
"metadataPartition": {
"available": 660270088192,
"total": 873862266880
}
}
]
}
}
}
}
```
#### GetClusterHealth `GET /v0/health`
#### GetClusterHealth `GET /v1/health`
Returns the cluster's current health in JSON format, with the following variables:
- `status`: one of `Healthy`, `Degraded` or `Unavailable`:
- Healthy: Garage node is connected to all storage nodes
- Degraded: Garage node is not connected to all storage nodes, but a quorum of write nodes is available for all partitions
- Unavailable: a quorum of write nodes is not available for some partitions
- `known_nodes`: the number of nodes this Garage node has had a TCP connection to since the daemon started
- `connected_nodes`: the nubmer of nodes this Garage node currently has an open connection to
- `storage_nodes`: the number of storage nodes currently registered in the cluster layout
- `storage_nodes_ok`: the number of storage nodes to which a connection is currently open
- `status`: one of `healthy`, `degraded` or `unavailable`:
- healthy: Garage node is connected to all storage nodes
- degraded: Garage node is not connected to all storage nodes, but a quorum of write nodes is available for all partitions
- unavailable: a quorum of write nodes is not available for some partitions
- `knownNodes`: the number of nodes this Garage node has had a TCP connection to since the daemon started
- `connectedNodes`: the nubmer of nodes this Garage node currently has an open connection to
- `storageNodes`: the number of storage nodes currently registered in the cluster layout
- `storageNodesOk`: the number of storage nodes to which a connection is currently open
- `partitions`: the total number of partitions of the data (currently always 256)
- `partitions_quorum`: the number of partitions for which a quorum of write nodes is available
- `partitions_all_ok`: the number of partitions for which we are connected to all storage nodes responsible of storing it
- `partitionsQuorum`: the number of partitions for which a quorum of write nodes is available
- `partitionsAllOk`: the number of partitions for which we are connected to all storage nodes responsible of storing it
Contrarily to `GET /health`, this endpoint always returns a 200 OK HTTP response code.
@ -157,18 +191,18 @@ Example response body:
```json
{
"status": "Degraded",
"known_nodes": 3,
"connected_nodes": 2,
"storage_nodes": 3,
"storage_nodes_ok": 2,
"status": "degraded",
"knownNodes": 3,
"connectedNodes": 3,
"storageNodes": 4,
"storageNodesOk": 3,
"partitions": 256,
"partitions_quorum": 256,
"partitions_all_ok": 0
"partitionsQuorum": 256,
"partitionsAllOk": 64
}
```
#### ConnectClusterNodes `POST /v0/connect`
#### ConnectClusterNodes `POST /v1/connect`
Instructs this Garage node to connect to other Garage nodes at specified addresses.
@ -198,7 +232,7 @@ Example response:
]
```
#### GetClusterLayout `GET /v0/layout`
#### GetClusterLayout `GET /v1/layout`
Returns the cluster's current layout in JSON, including:
@ -212,42 +246,54 @@ Example response body:
```json
{
"version": 12,
"roles": {
"ec79480e0ce52ae26fd00c9da684e4fa56658d9c64cdcecb094e936de0bfe71f": {
"roles": [
{
"id": "ec79480e0ce52ae26fd00c9da684e4fa56658d9c64cdcecb094e936de0bfe71f",
"zone": "dc1",
"capacity": 4,
"capacity": 10737418240,
"tags": [
"node1"
]
},
"4a6ae5a1d0d33bf895f5bb4f0a418b7dc94c47c0dd2eb108d1158f3c8f60b0ff": {
{
"id": "4a6ae5a1d0d33bf895f5bb4f0a418b7dc94c47c0dd2eb108d1158f3c8f60b0ff",
"zone": "dc1",
"capacity": 6,
"capacity": 10737418240,
"tags": [
"node2"
]
},
"23ffd0cdd375ebff573b20cc5cef38996b51c1a7d6dbcf2c6e619876e507cf27": {
{
"id": "23ffd0cdd375ebff573b20cc5cef38996b51c1a7d6dbcf2c6e619876e507cf27",
"zone": "dc2",
"capacity": 10,
"capacity": 10737418240,
"tags": [
"node3"
]
}
},
"stagedRoleChanges": {
"e2ee7984ee65b260682086ec70026165903c86e601a4a5a501c1900afe28d84b": {
],
"stagedRoleChanges": [
{
"id": "e2ee7984ee65b260682086ec70026165903c86e601a4a5a501c1900afe28d84b",
"remove": false,
"zone": "dc2",
"capacity": 5,
"capacity": 10737418240,
"tags": [
"node4"
]
}
{
"id": "23ffd0cdd375ebff573b20cc5cef38996b51c1a7d6dbcf2c6e619876e507cf27",
"remove": true,
"zone": null,
"capacity": null,
"tags": null,
}
]
}
```
#### UpdateClusterLayout `POST /v0/layout`
#### UpdateClusterLayout `POST /v1/layout`
Send modifications to the cluster layout. These modifications will
be included in the staged role changes, visible in subsequent calls
@ -259,8 +305,9 @@ the layout.
Request body format:
```json
{
<node_id>: {
[
{
"id": <node_id>,
"capacity": <new_capacity>,
"zone": <new_zone>,
"tags": [
@ -268,17 +315,22 @@ Request body format:
...
]
},
<node_id_to_remove>: null,
...
}
{
"id": <node_id_to_remove>,
"remove": true
}
]
```
Contrary to the CLI that may update only a subset of the fields
`capacity`, `zone` and `tags`, when calling this API all of these
values must be specified.
This returns the new cluster layout with the proposed staged changes,
as returned by GetClusterLayout.
#### ApplyClusterLayout `POST /v0/layout/apply`
#### ApplyClusterLayout `POST /v1/layout/apply`
Applies to the cluster the layout changes currently registered as
staged layout changes.
@ -295,7 +347,10 @@ Similarly to the CLI, the body must include the version of the new layout
that will be created, which MUST be 1 + the value of the currently
existing layout in the cluster.
#### RevertClusterLayout `POST /v0/layout/revert`
This returns the message describing all the calculations done to compute the new
layout, as well as the description of the layout as returned by GetClusterLayout.
#### RevertClusterLayout `POST /v1/layout/revert`
Clears all of the staged layout changes.
@ -313,10 +368,13 @@ Similarly to the CLI, the body must include the incremented
version number, which MUST be 1 + the value of the currently
existing layout in the cluster.
This returns the new cluster layout with all changes reverted,
as returned by GetClusterLayout.
### Access key operations
#### ListKeys `GET /v0/key`
#### ListKeys `GET /v1/key`
Returns all API access keys in the cluster.
@ -335,34 +393,8 @@ Example response:
]
```
#### CreateKey `POST /v0/key`
Creates a new API access key.
Request body format:
```json
{
"name": "NameOfMyKey"
}
```
#### ImportKey `POST /v0/key/import`
Imports an existing API key.
Request body format:
```json
{
"accessKeyId": "GK31c2f218a2e44f485b94239e",
"secretAccessKey": "b892c0665f0ada8a4755dae98baa3b133590e11dae3bcc1f9d769d67f16c3835",
"name": "NameOfMyKey"
}
```
#### GetKeyInfo `GET /v0/key?id=<acces key id>`
#### GetKeyInfo `GET /v0/key?search=<pattern>`
#### GetKeyInfo `GET /v1/key?id=<acces key id>`
#### GetKeyInfo `GET /v1/key?search=<pattern>`
Returns information about the requested API access key.
@ -370,6 +402,9 @@ If `id` is set, the key is looked up using its exact identifier (faster).
If `search` is set, the key is looked up using its name or prefix
of identifier (slower, all keys are enumerated to do this).
Optionnally, the query parameter `showSecretKey=true` can be set to reveal the
associated secret access key.
Example response:
```json
@ -433,11 +468,40 @@ Example response:
}
```
#### DeleteKey `DELETE /v0/key?id=<acces key id>`
#### CreateKey `POST /v1/key`
Deletes an API access key.
Creates a new API access key.
#### UpdateKey `POST /v0/key?id=<acces key id>`
Request body format:
```json
{
"name": "NameOfMyKey"
}
```
This returns the key info, including the created secret key,
in the same format as the result of GetKeyInfo.
#### ImportKey `POST /v1/key/import`
Imports an existing API key.
This will check that the imported key is in the valid format, i.e.
is a key that could have been generated by Garage.
Request body format:
```json
{
"accessKeyId": "GK31c2f218a2e44f485b94239e",
"secretAccessKey": "b892c0665f0ada8a4755dae98baa3b133590e11dae3bcc1f9d769d67f16c3835",
"name": "NameOfMyKey"
}
```
This returns the key info in the same format as the result of GetKeyInfo.
#### UpdateKey `POST /v1/key?id=<acces key id>`
Updates information about the specified API access key.
@ -453,14 +517,20 @@ Request body format:
}
```
All fields (`name`, `allow` and `deny`) are optionnal.
All fields (`name`, `allow` and `deny`) are optional.
If they are present, the corresponding modifications are applied to the key, otherwise nothing is changed.
The possible flags in `allow` and `deny` are: `createBucket`.
This returns the key info in the same format as the result of GetKeyInfo.
#### DeleteKey `DELETE /v1/key?id=<acces key id>`
Deletes an API access key.
### Bucket operations
#### ListBuckets `GET /v0/bucket`
#### ListBuckets `GET /v1/bucket`
Returns all storage buckets in the cluster.
@ -502,8 +572,8 @@ Example response:
]
```
#### GetBucketInfo `GET /v0/bucket?id=<bucket id>`
#### GetBucketInfo `GET /v0/bucket?globalAlias=<alias>`
#### GetBucketInfo `GET /v1/bucket?id=<bucket id>`
#### GetBucketInfo `GET /v1/bucket?globalAlias=<alias>`
Returns information about the requested storage bucket.
@ -535,7 +605,10 @@ Example response:
],
"objects": 14827,
"bytes": 13189855625,
"unfinshedUploads": 0,
"unfinishedUploads": 1,
"unfinishedMultipartUploads": 1,
"unfinishedMultipartUploadParts": 11,
"unfinishedMultipartUploadBytes": 41943040,
"quotas": {
"maxSize": null,
"maxObjects": null
@ -543,7 +616,7 @@ Example response:
}
```
#### CreateBucket `POST /v0/bucket`
#### CreateBucket `POST /v1/bucket`
Creates a new storage bucket.
@ -583,13 +656,7 @@ or no alias at all.
Technically, you can also specify both `globalAlias` and `localAlias` and that would create
two aliases, but I don't see why you would want to do that.
#### DeleteBucket `DELETE /v0/bucket?id=<bucket id>`
Deletes a storage bucket. A bucket cannot be deleted if it is not empty.
Warning: this will delete all aliases associated with the bucket!
#### UpdateBucket `PUT /v0/bucket?id=<bucket id>`
#### UpdateBucket `PUT /v1/bucket?id=<bucket id>`
Updates configuration of the given bucket.
@ -609,7 +676,7 @@ Request body format:
}
```
All fields (`websiteAccess` and `quotas`) are optionnal.
All fields (`websiteAccess` and `quotas`) are optional.
If they are present, the corresponding modifications are applied to the bucket, otherwise nothing is changed.
In `websiteAccess`: if `enabled` is `true`, `indexDocument` must be specified.
@ -621,9 +688,16 @@ In `quotas`: new values of `maxSize` and `maxObjects` must both be specified, or
to remove the quotas. An absent value will be considered the same as a `null`. It is not possible
to change only one of the two quotas.
#### DeleteBucket `DELETE /v1/bucket?id=<bucket id>`
Deletes a storage bucket. A bucket cannot be deleted if it is not empty.
Warning: this will delete all aliases associated with the bucket!
### Operations on permissions for keys on buckets
#### BucketAllowKey `POST /v0/bucket/allow`
#### BucketAllowKey `POST /v1/bucket/allow`
Allows a key to do read/write/owner operations on a bucket.
@ -644,7 +718,7 @@ Request body format:
Flags in `permissions` which have the value `true` will be activated.
Other flags will remain unchanged.
#### BucketDenyKey `POST /v0/bucket/deny`
#### BucketDenyKey `POST /v1/bucket/deny`
Denies a key from doing read/write/owner operations on a bucket.
@ -668,19 +742,19 @@ Other flags will remain unchanged.
### Operations on bucket aliases
#### GlobalAliasBucket `PUT /v0/bucket/alias/global?id=<bucket id>&alias=<global alias>`
#### GlobalAliasBucket `PUT /v1/bucket/alias/global?id=<bucket id>&alias=<global alias>`
Empty body. Creates a global alias for a bucket.
#### GlobalUnaliasBucket `DELETE /v0/bucket/alias/global?id=<bucket id>&alias=<global alias>`
#### GlobalUnaliasBucket `DELETE /v1/bucket/alias/global?id=<bucket id>&alias=<global alias>`
Removes a global alias for a bucket.
#### LocalAliasBucket `PUT /v0/bucket/alias/local?id=<bucket id>&accessKeyId=<access key ID>&alias=<local alias>`
#### LocalAliasBucket `PUT /v1/bucket/alias/local?id=<bucket id>&accessKeyId=<access key ID>&alias=<local alias>`
Empty body. Creates a local alias for a bucket in the namespace of a specific access key.
#### LocalUnaliasBucket `DELETE /v0/bucket/alias/local?id=<bucket id>&accessKeyId<access key ID>&alias=<local alias>`
#### LocalUnaliasBucket `DELETE /v1/bucket/alias/local?id=<bucket id>&accessKeyId<access key ID>&alias=<local alias>`
Removes a local alias for a bucket in the namespace of a specific access key.

View file

@ -146,7 +146,7 @@ in a bucket, as the partition key becomes the sort key in the index.
How indexing works:
- Each node keeps a local count of how many items it stores for each partition,
in a local Sled tree that is updated atomically when an item is modified.
in a local database tree that is updated atomically when an item is modified.
- These local counters are asynchronously stored in the index table which is
a regular Garage table spread in the network. Counters are stored as LWW values,
so basically the final table will have the following structure:

13
doc/optimal_layout_report/.gitignore vendored Normal file
View file

@ -0,0 +1,13 @@
optimal_layout.aux
optimal_layout.log
optimal_layout.synctex.gz
optimal_layout.bbl
optimal_layout.blg
geodistrib.aux
geodistrib.bbl
geodistrib.blg
geodistrib.log
geodistrib.out
geodistrib.synctex.gz

Binary file not shown.

File diff suppressed because it is too large Load diff

After

Width:  |  Height:  |  Size: 161 KiB

Binary file not shown.

File diff suppressed because it is too large Load diff

After

Width:  |  Height:  |  Size: 560 KiB

Binary file not shown.

File diff suppressed because it is too large Load diff

After

Width:  |  Height:  |  Size: 287 KiB

Binary file not shown.

File diff suppressed because it is too large Load diff

After

Width:  |  Height:  |  Size: 112 KiB

Binary file not shown.

File diff suppressed because it is too large Load diff

After

Width:  |  Height:  |  Size: 270 KiB

Binary file not shown.

View file

@ -0,0 +1,317 @@
\documentclass[]{article}
\usepackage{amsmath,amssymb}
\usepackage{amsthm}
\usepackage{stmaryrd}
\usepackage{graphicx,xcolor}
\usepackage{hyperref}
\usepackage{algorithm,algpseudocode,float}
\renewcommand\thesubsubsection{\Alph{subsubsection})}
\newtheorem{proposition}{Proposition}
%opening
\title{An algorithm for geo-distributed and redundant storage in Garage}
\author{Mendes Oulamara \\ \emph{mendes@deuxfleurs.fr}}
\date{}
\begin{document}
\maketitle
\begin{abstract}
Garage
\end{abstract}
\section{Introduction}
Garage\footnote{\url{https://garagehq.deuxfleurs.fr/}} is an open-source distributed object storage service tailored for self-hosting. It was designed by the Deuxfleurs association\footnote{\url{https://deuxfleurs.fr/}} to enable small structures (associations, collectives, small companies) to share storage resources to reliably self-host their data, possibly with old and non-reliable machines.
To achieve these reliability and availability goals, the data is broken into \emph{partitions} and every partition is replicated over 3 different machines (that we call \emph{nodes}). When the data is queried, a consensus algorithm allows to fetch it from one of the nodes. A \emph{replication factor} of 3 ensures the best guarantees in the consensus algorithm \cite{ADD RREF}, but this parameter can be different.
Moreover, if the nodes are spread over different \emph{zones} (different houses, offices, cities\dots), we can ask the data to be replicated over nodes belonging to different zones, to improve the storage robustness against zone failure (such as power outage). To do so, we set a \emph{redundancy parameter}, that is no more than the replication factor, and we ask that any partition is replicated over this number of zones at least.
In this work, we propose a repartition algorithm that, given the nodes specifications and the replication and redundancy parameters, computes an optimal assignation of partitions to nodes. We say that the assignation is optimal in the sense that it maximizes the size of the partitions, and hence the effective storage capacity of the system.
Moreover, when a former assignation exists, which is not optimal anymore due to nodes or zones updates, our algorithm computes a new optimal assignation that minimizes the amount of data to be transferred during the assignation update (the \emph{transfer load}).
We call the set of nodes cooperating to store the data a \emph{cluster}, and a description of the nodes, zones and the assignation of partitions to nodes a \emph{cluster layout}
\subsection{Notations}
Let $k$ be some fixed parameter value, typically 8, that we call the ``partition bits''.
Every object to be stored in the system is split into data blocks of fixed size. We compute a hash $h(\mathbf{b})$ of every such block $\mathbf{b}$, and we define the $k$ last bits of this hash to be the partition number $p(\mathbf{b})$ of the block. This label can take $P=2^k$ different values, and hence there are $P$ different partitions. We denote $\mathbf{P}$ the set of partition labels (i.e. $\mathbf{P}=\llbracket1,P\rrbracket$).
We are given a set $\mathbf{N}$ of $N$ nodes and a set $\mathbf{Z}$ of $Z$ zones. Every node $n$ has a non-negative storage capacity $c_n\ge 0$ and belongs to a zone $z_n\in \mathbf{Z}$. We are also given a replication parameter $\rho_\mathbf{N}$ and a redundancy parameter $\rho_\mathbf{Z}$ such that $1\le \rho_\mathbf{Z} \le \rho_\mathbf{N}$ (typical values would be $\rho_N=3$ and $\rho_Z=2$).
Our goal is to compute an assignment $\alpha = (\alpha_p^1, \ldots, \alpha_p^{\rho_\mathbf{N}})_{p\in \mathbf{P}}$ such that every partition $p$ is associated to $\rho_\mathbf{N}$ distinct nodes $\alpha_p^1, \ldots, \alpha_p^{\rho_\mathbf{N}} \in \mathbf{N}$ and these nodes belong to at least $\rho_\mathbf{Z}$ distinct zones. Among the possible assignations, we choose one that \emph{maximizes} the effective storage capacity of the cluster. If the layout contained a previous assignment $\alpha'$, we \emph{minimize} the amount of data to transfer during the layout update by making $\alpha$ as close as possible to $\alpha'$. These maximization and minimization are described more formally in the following section.
\subsection{Optimization parameters}
To link the effective storage capacity of the cluster to partition assignment, we make the following assumption:
\begin{equation}
\tag{H1}
\text{\emph{All partitions have the same size $s$.}}
\end{equation}
This assumption is justified by the dispersion of the hashing function, when the number of partitions is small relative to the number of stored blocks.
Every node $n$ wille store some number $p_n$ of partitions (it is the number of partitions $p$ such that $n$ appears in the $\alpha_p$). Hence the partitions stored by $n$ (and hence all partitions by our assumption) have there size bounded by $c_n/p_n$. This remark leads us to define the optimal size that we will want to maximize:
\begin{equation}
\label{eq:optimal}
\tag{OPT}
s^* = \min_{n \in N} \frac{c_n}{p_n}.
\end{equation}
When the capacities of the nodes are updated (this includes adding or removing a node), we want to update the assignment as well. However, transferring the data between nodes has a cost and we would like to limit the number of changes in the assignment. We make the following assumption:
\begin{equation}
\tag{H2}
\text{\emph{Nodes updates happen rarely relatively to block operations.}}
\end{equation}
This assumption justifies that when we compute the new assignment $\alpha$, it is worth to optimize the partition size \eqref{eq:optimal} first, and then, among the possible optimal solution, to try to minimize the number of partition transfers. More formally, we minimize the distance between two assignments defined by
\begin{equation}
d(\alpha, \alpha') := \#\{ (n,p) \in \mathbf{N}\times\mathbf{P} ~|~ n\in \alpha_p \triangle \alpha'_p \}
\end{equation}
where the symmetric difference $\alpha_p \triangle \alpha'_p$ denotes the nodes appearing in one of the assignations but not in both.
\section{Computation of an optimal assignment}
The algorithm that we propose takes as inputs the cluster layout parameters $\mathbf{N}$, $\mathbf{Z}$, $\mathbf{P}$, $(c_n)_{n\in \mathbf{N}}$, $\rho_\mathbf{N}$, $\rho_\mathbf{Z}$, that we defined in the introduction, together with the former assignation $\alpha'$ (if any). The computation of the new optimal assignation $\alpha^*$ is done in three successive steps that will be detailed in the following sections. The first step computes the largest partition size $s^*$ that an assignation can achieve. The second step computes an optimal candidate assignment $\alpha$ that achieves $s^*$ and a heuristic is used in the computation to make it hopefully close to $\alpha'$. The third steps modifies $\alpha$ iteratively to reduces $d(\alpha, \alpha')$ and yields an assignation $\alpha^*$ achieving $s^*$, and minimizing $d(\cdot, \alpha')$ among such assignations.
We will explain in the next section how to represent an assignment $\alpha$ by a flow $f$ on a weighted graph $G$ to enable the use of flow and graph algorithms. The main function of the algorithm can be written as follows.
\subsubsection*{Algorithm}
\begin{algorithmic}[1]
\Function{Compute Layout}{$\mathbf{N}$, $\mathbf{Z}$, $\mathbf{P}$, $(c_n)_{n\in \mathbf{N}}$, $\rho_\mathbf{N}$, $\rho_\mathbf{Z}$, $\alpha'$}
\State $s^* \leftarrow$ \Call{Compute Partition Size}{$\mathbf{N}$, $\mathbf{Z}$, $\mathbf{P}$, $(c_n)_{n\in \mathbf{N}}$, $\rho_\mathbf{N}$, $\rho_\mathbf{Z}$}
\State $G \leftarrow G(s^*)$
\State $f \leftarrow$ \Call{Compute Candidate Assignment}{$G$, $\alpha'$}
\State $f^* \leftarrow$ \Call{Minimize transfer load}{$G$, $f$, $\alpha'$}
\State Build $\alpha^*$ from $f^*$
\State \Return $\alpha^*$
\EndFunction
\end{algorithmic}
\subsubsection*{Complexity}
As we will see in the next sections, the worst case complexity of this algorithm is $O(P^2 N^2)$. The minimization of transfer load is the most expensive step, and it can run with a timeout since it is only an optimization step. Without this step (or with a smart timeout), the worst cas complexity can be $O((PN)^{3/2}\log C)$ where $C$ is the total storage capacity of the cluster.
\subsection{Determination of the partition size $s^*$}
We will represent an assignment $\alpha$ as a flow in a specific graph $G$. We will not compute the optimal partition size $s^*$ a priori, but we will determine it by dichotomy, as the largest size $s$ such that the maximal flow achievable on $G=G(s)$ has value $\rho_\mathbf{N}P$. We will assume that the capacities are given in a small enough unit (say, Megabytes), and we will determine $s^*$ at the precision of the given unit.
Given some candidate size value $s$, we describe the oriented weighted graph $G=(V,E)$ with vertex set $V$ arc set $E$ (see Figure \ref{fig:flowgraph}).
The set of vertices $V$ contains the source $\mathbf{s}$, the sink $\mathbf{t}$, vertices
$\mathbf{p^+, p^-}$ for every partition $p$, vertices $\mathbf{x}_{p,z}$ for every partition $p$ and zone $z$, and vertices $\mathbf{n}$ for every node $n$.
The set of arcs $E$ contains:
\begin{itemize}
\item ($\mathbf{s}$,$\mathbf{p}^+$, $\rho_\mathbf{Z}$) for every partition $p$;
\item ($\mathbf{s}$,$\mathbf{p}^-$, $\rho_\mathbf{N}-\rho_\mathbf{Z}$) for every partition $p$;
\item ($\mathbf{p}^+$,$\mathbf{x}_{p,z}$, 1) for every partition $p$ and zone $z$;
\item ($\mathbf{p}^-$,$\mathbf{x}_{p,z}$, $\rho_\mathbf{N}-\rho_\mathbf{Z}$) for every partition $p$ and zone $z$;
\item ($\mathbf{x}_{p,z}$,$\mathbf{n}$, 1) for every partition $p$, zone $z$ and node $n\in z$;
\item ($\mathbf{n}$, $\mathbf{t}$, $\lfloor c_n/s \rfloor$) for every node $n$.
\end{itemize}
\begin{figure}
\centering
\includegraphics[width=\linewidth]{figures/flow_graph_param}
\caption{An example of graph $G(s)$. Arcs are oriented from left to right, and unlabeled arcs have capacity 1. In this example, nodes $n_1,n_2,n_3$ belong to zone $z_1$, and nodes $n_4,n_5$ belong to zone $z_2$.}
\label{fig:flowgraph}
\end{figure}
In the following complexity calculations, we will use the number of vertices and edges of $G$. Remark from now that $\# V = O(PZ)$ and $\# E = O(PN)$.
\begin{proposition}
An assignment $\alpha$ is realizable with partition size $s$ and the redundancy constraints $(\rho_\mathbf{N},\rho_\mathbf{Z})$ if and only if there exists a maximal flow function $f$ in $G$ with total flow $\rho_\mathbf{N}P$, such that the arcs ($\mathbf{x}_{p,z}$,$\mathbf{n}$, 1) used are exactly those for which $p$ is associated to $n$ in $\alpha$.
\end{proposition}
\begin{proof}
Given such flow $f$, we can reconstruct a candidate $\alpha$. In $f$, the flow passing through $\mathbf{p^+}$ and $\mathbf{p^-}$ is $\rho_\mathbf{N}$, and since the outgoing capacity of every $\mathbf{x}_{p,z}$ is 1, every partition is associated to $\rho_\mathbf{N}$ distinct nodes. The fraction $\rho_\mathbf{Z}$ of the flow passing through every $\mathbf{p^+}$ must be spread over as many distinct zones as every arc outgoing from $\mathbf{p^+}$ has capacity 1. So the reconstructed $\alpha$ verifies the redundancy constraints. For every node $n$, the flow between $\mathbf{n}$ and $\mathbf{t}$ corresponds to the number of partitions associated to $n$. By construction of $f$, this does not exceed $\lfloor c_n/s \rfloor$. We assumed that the partition size is $s$, hence this association does not exceed the storage capacity of the nodes.
In the other direction, given an assignment $\alpha$, one can similarly check that the facts that $\alpha$ respects the redundancy constraints, and the storage capacities of the nodes, are necessary condition to construct a maximal flow function $f$.
\end{proof}
\textbf{Implementation remark:} In the flow algorithm, while exploring the graph, we explore the neighbours of every vertex in a random order to heuristically spread the associations between nodes and partitions.
\subsubsection*{Algorithm}
With this result mind, we can describe the first step of our algorithm. All divisions are supposed to be integer divisions.
\begin{algorithmic}[1]
\Function{Compute Partition Size}{$\mathbf{N}$, $\mathbf{Z}$, $\mathbf{P}$, $(c_n)_{n\in \mathbf{N}}$, $\rho_\mathbf{N}$, $\rho_\mathbf{Z}$}
\State Build the graph $G=G(s=1)$
\State $ f \leftarrow$ \Call{Maximal flow}{$G$}
\If{$f.\mathrm{total flow} < \rho_\mathbf{N}P$}
\State \Return Error: capacities too small or constraints too strong.
\EndIf
\State $s^- \leftarrow 1$
\State $s^+ \leftarrow 1+\frac{1}{\rho_\mathbf{N}}\sum_{n \in \mathbf{N}} c_n$
\While{$s^-+1 < s^+$}
\State Build the graph $G=G(s=(s^-+s^+)/2)$
\State $ f \leftarrow$ \Call{Maximal flow}{$G$}
\If{$f.\mathrm{total flow} < \rho_\mathbf{N}P$}
\State $s^+ \leftarrow (s^- + s^+)/2$
\Else
\State $s^- \leftarrow (s^- + s^+)/2$
\EndIf
\EndWhile
\State \Return $s^-$
\EndFunction
\end{algorithmic}
\subsubsection*{Complexity}
To compute the maximal flow, we use Dinic's algorithm. Its complexity on general graphs is $O(\#V^2 \#E)$, but on graphs with edge capacity bounded by a constant, it turns out to be $O(\#E^{3/2})$. The graph $G$ does not fall in this case since the capacities of the arcs incoming to $\mathbf{t}$ are far from bounded. However, the proof of this complexity function works readily for graphs where we only ask the edges \emph{not} incoming to the sink $\mathbf{t}$ to have their capacities bounded by a constant. One can find the proof of this claim in \cite[Section 2]{even1975network}.
The dichotomy adds a logarithmic factor $\log (C)$ where $C=\sum_{n \in \mathbf{N}} c_n$ is the total capacity of the cluster. The total complexity of this first function is hence
$O(\#E^{3/2}\log C ) = O\big((PN)^{3/2} \log C\big)$.
\subsubsection*{Metrics}
We can display the discrepancy between the computed $s^*$ and the best size we could have hoped for the given total capacity, that is $C/\rho_\mathbf{N}$.
\subsection{Computation of a candidate assignment}
Now that we have the optimal partition size $s^*$, to compute a candidate assignment it would be enough to compute a maximal flow function $f$ on $G(s^*)$. This is what we do if there is no former assignation $\alpha'$.
If there is some $\alpha'$, we add a step that will heuristically help to obtain a candidate $\alpha$ closer to $\alpha'$. We fist compute a flow function $\tilde{f}$ that uses only the partition-to-node associations appearing in $\alpha'$. Most likely, $\tilde{f}$ will not be a maximal flow of $G(s^*)$. In Dinic's algorithm, we can start from a non maximal flow function and then discover improving paths. This is what we do by starting from $\tilde{f}$. The hope\footnote{This is only a hope, because one can find examples where the construction of $f$ from $\tilde{f}$ produces an assignment $\alpha$ that is not as close as possible to $\alpha'$.} is that the final flow function $f$ will tend to keep the associations appearing in $\tilde{f}$.
More formally, we construct the graph $G_{|\alpha'}$ from $G$ by removing all the arcs $(\mathbf{x}_{p,z},\mathbf{n}, 1)$ where $p$ is not associated to $n$ in $\alpha'$. We compute a maximal flow function $\tilde{f}$ in $G_{|\alpha'}$. The flow $\tilde{f}$ is also a valid (most likely non maximal) flow function on $G$. We compute a maximal flow function $f$ on $G$ by starting Dinic's algorithm on $\tilde{f}$.
\subsubsection*{Algorithm}
\begin{algorithmic}[1]
\Function{Compute Candidate Assignment}{$G$, $\alpha'$}
\State Build the graph $G_{|\alpha'}$
\State $ \tilde{f} \leftarrow$ \Call{Maximal flow}{$G_{|\alpha'}$}
\State $ f \leftarrow$ \Call{Maximal flow from flow}{$G$, $\tilde{f}$}
\State \Return $f$
\EndFunction
\end{algorithmic}
~
\textbf{Remark:} The function ``Maximal flow'' can be just seen as the function ``Maximal flow from flow'' called with the zero flow function as starting flow.
\subsubsection*{Complexity}
With the considerations of the last section, we have the complexity of the Dinic's algorithm $O(\#E^{3/2}) = O((PN)^{3/2})$.
\subsubsection*{Metrics}
We can display the flow value of $\tilde{f}$, which is an upper bound of the distance between $\alpha$ and $\alpha'$. It might be more a Debug level display than Info.
\subsection{Minimization of the transfer load}
Now that we have a candidate flow function $f$, we want to modify it to make its corresponding assignation $\alpha$ as close as possible to $\alpha'$. Denote by $f'$ the maximal flow corresponding to $\alpha'$, and let $d(f, \alpha')=d(f, f'):=d(\alpha,\alpha')$\footnote{It is the number of arcs of type $(\mathbf{x}_{p,z},\mathbf{n})$ saturated in one flow and not in the other.}.
We want to build a sequence $f=f_0, f_1, f_2 \dots$ of maximal flows such that $d(f_i, \alpha')$ decreases as $i$ increases. The distance being a non-negative integer, this sequence of flow functions must be finite. We now explain how to find some improving $f_{i+1}$ from $f_i$.
For any maximal flow $f$ in $G$, we define the oriented weighted graph $G_f=(V, E_f)$ as follows. The vertices of $G_f$ are the same as the vertices of $G$. $E_f$ contains the arc $(v_1,v_2, w)$ between vertices $v_1,v_2\in V$ with weight $w$ if and only if the arc $(v_1,v_2)$ is not saturated in $f$ (i.e. $c(v_1,v_2)-f(v_1,v_2) \ge 1$, we also consider reversed arcs). The weight $w$ is:
\begin{itemize}
\item $-1$ if $(v_1,v_2)$ is of type $(\mathbf{x}_{p,z},\mathbf{n})$ or $(\mathbf{x}_{p,z},\mathbf{n})$ and is saturated in only one of the two flows $f,f'$;
\item $+1$ if $(v_1,v_2)$ is of type $(\mathbf{x}_{p,z},\mathbf{n})$ or $(\mathbf{x}_{p,z},\mathbf{n})$ and is saturated in either both or none of the two flows $f,f'$;
\item $0$ otherwise.
\end{itemize}
If $\gamma$ is a simple cycle of arcs in $G_f$, we define its weight $w(\gamma)$ as the sum of the weights of its arcs. We can add $+1$ to the value of $f$ on the arcs of $\gamma$, and by construction of $G_f$ and the fact that $\gamma$ is a cycle, the function that we get is still a valid flow function on $G$, it is maximal as it has the same flow value as $f$. We denote this new function $f+\gamma$.
\begin{proposition}
Given a maximal flow $f$ and a simple cycle $\gamma$ in $G_f$, we have $d(f+\gamma, f') - d(f,f') = w(\gamma)$.
\end{proposition}
\begin{proof}
Let $X$ be the set of arcs of type $(\mathbf{x}_{p,z},\mathbf{n})$. Then we can express $d(f,f')$ as
\begin{align*}
d(f,f') & = \#\{e\in X ~|~ f(e)\neq f'(e)\}
= \sum_{e\in X} 1_{f(e)\neq f'(e)} \\
& = \frac{1}{2}\big( \#X + \sum_{e\in X} 1_{f(e)\neq f'(e)} - 1_{f(e)= f'(e)} \big).
\end{align*}
We can express the cycle weight as
\begin{align*}
w(\gamma) & = \sum_{e\in X, e\in \gamma} - 1_{f(e)\neq f'(e)} + 1_{f(e)= f'(e)}.
\end{align*}
Remark that since we passed on unit of flow in $\gamma$ to construct $f+\gamma$, we have for any $e\in X$, $f(e)=f'(e)$ if and only if $(f+\gamma)(e) \neq f'(e)$.
Hence
\begin{align*}
w(\gamma) & = \frac{1}{2}(w(\gamma) + w(\gamma)) \\
&= \frac{1}{2} \Big(
\sum_{e\in X, e\in \gamma} - 1_{f(e)\neq f'(e)} + 1_{f(e)= f'(e)} \\
& \qquad +
\sum_{e\in X, e\in \gamma} 1_{(f+\gamma)(e)\neq f'(e)} + 1_{(f+\gamma)(e)= f'(e)}
\Big).
\end{align*}
Plugging this in the previous equation, we find that
$$d(f,f')+w(\gamma) = d(f+\gamma, f').$$
\end{proof}
This result suggests that given some flow $f_i$, we just need to find a negative cycle $\gamma$ in $G_{f_i}$ to construct $f_{i+1}$ as $f_i+\gamma$. The following proposition ensures that this greedy strategy reaches an optimal flow.
\begin{proposition}
For any maximal flow $f$, $G_f$ contains a negative cycle if and only if there exists a maximal flow $f^*$ in $G$ such that $d(f^*, f') < d(f, f')$.
\end{proposition}
\begin{proof}
Suppose that there is such flow $f^*$. Define the oriented multigraph $M_{f,f^*}=(V,E_M)$ with the same vertex set $V$ as in $G$, and for every $v_1,v_2 \in V$, $E_M$ contains $(f^*(v_1,v_2) - f(v_1,v_2))_+$ copies of the arc $(v_1,v_2)$. For every vertex $v$, its total degree (meaning its outer degree minus its inner degree) is equal to
\begin{align*}
\deg v & = \sum_{u\in V} (f^*(v,u) - f(v,u))_+ - \sum_{u\in V} (f^*(u,v) - f(u,v))_+ \\
& = \sum_{u\in V} f^*(v,u) - f(v,u) = \sum_{u\in V} f^*(v,u) - \sum_{u\in V} f(v,u).
\end{align*}
The last two sums are zero for any inner vertex since $f,f^*$ are flows, and they are equal on the source and sink since the two flows are both maximal and have hence the same value. Thus, $\deg v = 0$ for every vertex $v$.
This implies that the multigraph $M_{f,f^*}$ is the union of disjoint simple cycles. $f$ can be transformed into $f^*$ by pushing a mass 1 along all these cycles in any order. Since $d(f^*, f')<d(f,f')$, there must exists one of these simple cycles $\gamma$ with $d(f+\gamma, f') < d(f, f')$. Finally, since we can push a mass in $f$ along $\gamma$, it must appear in $G_f$. Hence $\gamma$ is a cycle of $G_f$ with negative weight.
\end{proof}
In the next section we describe the corresponding algorithm. Instead of discovering only one cycle, we are allowed to discover a set $\Gamma$ of disjoint negative cycles.
\subsubsection*{Algorithm}
\begin{algorithmic}[1]
\Function{Minimize transfer load}{$G$, $f$, $\alpha'$}
\State Build the graph $G_f$
\State $\Gamma \leftarrow$ \Call{Detect Negative Cycles}{$G_f$}
\While{$\Gamma \neq \emptyset$}
\ForAll{$\gamma \in \Gamma$}
\State $f \leftarrow f+\gamma$
\EndFor
\State Update $G_f$
\State $\Gamma \leftarrow$ \Call{Detect Negative Cycles}{$G_f$}
\EndWhile
\State \Return $f$
\EndFunction
\end{algorithmic}
\subsubsection*{Complexity}
The distance $d(f,f')$ is bounded by the maximal number of differences in the associated assignment. If these assignment are totally disjoint, this distance is $2\rho_N P$. At every iteration of the While loop, the distance decreases, so there is at most $O(\rho_N P) = O(P)$ iterations.
The detection of negative cycle is done with the Bellman-Ford algorithm, whose complexity should normally be $O(\#E\#V)$. In our case, it amounts to $O(P^2ZN)$. Multiplied by the complexity of the outer loop, it amounts to $O(P^3ZN)$ which is a lot when the number of partitions and nodes starts to be large. To avoid that, we adapt the Bellman-Ford algorithm.
The Bellman-Ford algorithm runs $\#V$ iterations of an outer loop, and an inner loop over $E$. The idea is to compute the shortest paths from a source vertex $v$ to all other vertices. After $k$ iterations of the outer loop, the algorithm has computed all shortest path of length at most $k$. All simple paths have length at most $\#V-1$, so if there is an update in the last iteration of the loop, it means that there is a negative cycle in the graph. The observation that will enable us to improve the complexity is the following:
\begin{proposition}
In the graph $G_f$ (and $G$), all simple paths have a length at most $4N$.
\end{proposition}
\begin{proof}
Since $f$ is a maximal flow, there is no outgoing edge from $\mathbf{s}$ in $G_f$. One can thus check than any simple path of length 4 must contain at least two node of type $\mathbf{n}$. Hence on a path, at most 4 arcs separate two successive nodes of type $\mathbf{n}$.
\end{proof}
Thus, in the absence of negative cycles, shortest paths in $G_f$ have length at most $4N$. So we can do only $4N+1$ iterations of the outer loop in the Bellman-Ford algorithm. This makes the complexity of the detection of one set of cycle to be $O(N\#E) = O(N^2 P)$.
With this improvement, the complexity of the whole algorithm is, in the worst case, $O(N^2P^2)$. However, since we detect several cycles at once and we start with a flow that might be close to the previous one, the number of iterations of the outer loop might be smaller in practice.
\subsubsection*{Metrics}
We can display the node and zone utilization ratio, by dividing the flow passing through them divided by their outgoing capacity. In particular, we can pinpoint saturated nodes and zones (i.e. used at their full potential).
We can display the distance to the previous assignment, and the number of partition transfers.
\bibliography{optimal_layout}
\bibliographystyle{ieeetr}
\end{document}

View file

@ -0,0 +1,11 @@
@article{even1975network,
title={Network flow and testing graph connectivity},
author={Even, Shimon and Tarjan, R Endre},
journal={SIAM journal on computing},
volume={4},
number={4},
pages={507--518},
year={1975},
publisher={SIAM}
}

Binary file not shown.

View file

@ -0,0 +1,709 @@
\documentclass[]{article}
\usepackage{amsmath,amssymb}
\usepackage{amsthm}
\usepackage{graphicx,xcolor}
\usepackage{algorithm,algpseudocode,float}
\renewcommand\thesubsubsection{\Alph{subsubsection})}
\newtheorem{proposition}{Proposition}
%opening
\title{Optimal partition assignment in Garage}
\author{Mendes}
\begin{document}
\maketitle
\section{Introduction}
\subsection{Context}
Garage is an open-source distributed storage service blablabla$\dots$
Every object to be stored in the system falls in a partition given by the last $k$ bits of its hash. There are $P=2^k$ partitions. Every partition will be stored on distinct nodes of the system. The goal of the assignment of partitions to nodes is to ensure (nodes and zone) redundancy and to be as efficient as possible.
\subsection{Formal description of the problem}
We are given a set of nodes $\mathbf{N}$ and a set of zones $\mathbf{Z}$. Every node $n$ has a non-negative storage capacity $c_n\ge 0$ and belongs to a zone $z\in \mathbf{Z}$. We are also given a number of partition $P>0$ (typically $P=256$).
We would like to compute an assignment of nodes to partitions. We will impose some redundancy constraints to this assignment, and under these constraints, we want our system to have the largest storage capacity possible. To link storage capacity to partition assignment, we make the following assumption:
\begin{equation}
\tag{H1}
\text{\emph{All partitions have the same size $s$.}}
\end{equation}
This assumption is justified by the dispersion of the hashing function, when the number of partitions is small relative to the number of stored large objects.
Every node $n$ wille store some number $k_n$ of partitions. Hence the partitions stored by $n$ (and hence all partitions by our assumption) have there size bounded by $c_n/k_n$. This remark leads us to define the optimal size that we will want to maximize:
\begin{equation}
\label{eq:optimal}
\tag{OPT}
s^* = \min_{n \in N} \frac{c_n}{k_n}.
\end{equation}
When the capacities of the nodes are updated (this includes adding or removing a node), we want to update the assignment as well. However, transferring the data between nodes has a cost and we would like to limit the number of changes in the assignment. We make the following assumption:
\begin{equation}
\tag{H2}
\text{\emph{Updates of capacity happens rarely relatively to object storing.}}
\end{equation}
This assumption justifies that when we compute the new assignment, it is worth to optimize the partition size \eqref{eq:optimal} first, and then, among the possible optimal solution, to try to minimize the number of partition transfers.
For now, in the following, we ask the following redundancy constraint:
\textbf{Parametric node and zone redundancy:} Given two integer parameters $1\le \rho_\mathbf{Z} \le \rho_\mathbf{N}$, we ask every partition to be stored on $\rho_\mathbf{N}$ distinct nodes, and these nodes must belong to at least $\rho_\mathbf{Z}$ distinct zones.
\textbf{Mode 3-strict:} every partition needs to be assignated to three nodes belonging to three different zones.
\textbf{Mode 3:} every partition needs to be assignated to three nodes. We try to spread the three nodes over different zones as much as possible.
\textbf{Warning:} This is a working document written incrementaly. The last version of the algorithm is the \textbf{parametric assignment} described in the next section.
\section{Computation of a parametric assignment}
\textbf{Attention : }We change notations in this section.
Notations : let $P$ be the number of partitions, $N$ the number of nodes, $Z$ the number of zones. Let $\mathbf{P,N,Z}$ be the label sets of, respectively, partitions, nodes and zones.
Let $s^*$ be the largest partition size achievable with the redundancy constraints. Let $(c_n)_{n\in \mathbf{N}}$ be the storage capacity of every node.
In this section, we propose a third specification of the problem. The user inputs two redundancy parameters $1\le \rho_\mathbf{Z} \le \rho_\mathbf{N}$. We compute an assignment $\alpha = (\alpha_p^1, \ldots, \alpha_p^{\rho_\mathbf{N}})_{p\in \mathbf{P}}$ such that every partition $p$ is associated to $\rho_\mathbf{N}$ distinct nodes $\alpha_p^1, \ldots, \alpha_p^{\rho_\mathbf{N}}$ and these nodes belong to at least $\rho_\mathbf{Z}$ distinct zones.
If the layout contained a previous assignment $\alpha'$, we try to minimize the amount of data to transfer during the layout update by making $\alpha$ as close as possible to $\alpha'$.
In the following subsections, we describe the successive steps of the algorithm we propose to compute $\alpha$.
\subsubsection*{Algorithm}
\begin{algorithmic}[1]
\Function{Compute Layout}{$\mathbf{N}$, $\mathbf{Z}$, $\mathbf{P}$, $(c_n)_{n\in \mathbf{N}}$, $\rho_\mathbf{N}$, $\rho_\mathbf{Z}$, $\alpha'$}
\State $s^* \leftarrow$ \Call{Compute Partition Size}{$\mathbf{N}$, $\mathbf{Z}$, $\mathbf{P}$, $(c_n)_{n\in \mathbf{N}}$, $\rho_\mathbf{N}$, $\rho_\mathbf{Z}$}
\State $G \leftarrow G(s^*)$
\State $f \leftarrow$ \Call{Compute Candidate Assignment}{$G$, $\alpha'$}
\State $f^* \leftarrow$ \Call{Minimize transfer load}{$G$, $f$, $\alpha'$}
\State Build $\alpha^*$ from $f^*$
\State \Return $\alpha^*$
\EndFunction
\end{algorithmic}
\subsubsection*{Complexity}
As we will see in the next sections, the worst case complexity of this algorithm is $O(P^2 N^2)$. The minimization of transfer load is the most expensive step, and it can run with a timeout since it is only an optimization step. Without this step (or with a smart timeout), the worst cas complexity can be $O((PN)^{3/2}\log C)$ where $C$ is the total storage capacity of the cluster.
\subsection{Determination of the partition size $s^*$}
Again, we will represent an assignment $\alpha$ as a flow in a specific graph $G$. We will not compute the optimal partition size $s^*$ a priori, but we will determine it by dichotomy, as the largest size $s$ such that the maximal flow achievable on $G=G(s)$ has value $\rho_\mathbf{N}P$. We will assume that the capacities are given in a small enough unit (say, Megabytes), and we will determine $s^*$ at the precision of the given unit.
Given some candidate size value $s$, we describe the oriented weighted graph $G=(V,E)$ with vertex set $V$ arc set $E$.
The set of vertices $V$ contains the source $\mathbf{s}$, the sink $\mathbf{t}$, vertices
$\mathbf{p^+, p^-}$ for every partition $p$, vertices $\mathbf{x}_{p,z}$ for every partition $p$ and zone $z$, and vertices $\mathbf{n}$ for every node $n$.
The set of arcs $E$ contains:
\begin{itemize}
\item ($\mathbf{s}$,$\mathbf{p}^+$, $\rho_\mathbf{Z}$) for every partition $p$;
\item ($\mathbf{s}$,$\mathbf{p}^-$, $\rho_\mathbf{N}-\rho_\mathbf{Z}$) for every partition $p$;
\item ($\mathbf{p}^+$,$\mathbf{x}_{p,z}$, 1) for every partition $p$ and zone $z$;
\item ($\mathbf{p}^-$,$\mathbf{x}_{p,z}$, $\rho_\mathbf{N}-\rho_\mathbf{Z}$) for every partition $p$ and zone $z$;
\item ($\mathbf{x}_{p,z}$,$\mathbf{n}$, 1) for every partition $p$, zone $z$ and node $n\in z$;
\item ($\mathbf{n}$, $\mathbf{t}$, $\lfloor c_n/s \rfloor$) for every node $n$.
\end{itemize}
In the following complexity calculations, we will use the number of vertices and edges of $G$. Remark from now that $\# V = O(PZ)$ and $\# E = O(PN)$.
\begin{proposition}
An assignment $\alpha$ is realizable with partition size $s$ and the redundancy constraints $(\rho_\mathbf{N},\rho_\mathbf{Z})$ if and only if there exists a maximal flow function $f$ in $G$ with total flow $\rho_\mathbf{N}P$, such that the arcs ($\mathbf{x}_{p,z}$,$\mathbf{n}$, 1) used are exactly those for which $p$ is associated to $n$ in $\alpha$.
\end{proposition}
\begin{proof}
Given such flow $f$, we can reconstruct a candidate $\alpha$. In $f$, the flow passing through $\mathbf{p^+}$ and $\mathbf{p^-}$ is $\rho_\mathbf{N}$, and since the outgoing capacity of every $\mathbf{x}_{p,z}$ is 1, every partition is associated to $\rho_\mathbf{N}$ distinct nodes. The fraction $\rho_\mathbf{Z}$ of the flow passing through every $\mathbf{p^+}$ must be spread over as many distinct zones as every arc outgoing from $\mathbf{p^+}$ has capacity 1. So the reconstructed $\alpha$ verifies the redundancy constraints. For every node $n$, the flow between $\mathbf{n}$ and $\mathbf{t}$ corresponds to the number of partitions associated to $n$. By construction of $f$, this does not exceed $\lfloor c_n/s \rfloor$. We assumed that the partition size is $s$, hence this association does not exceed the storage capacity of the nodes.
In the other direction, given an assignment $\alpha$, one can similarly check that the facts that $\alpha$ respects the redundancy constraints, and the storage capacities of the nodes, are necessary condition to construct a maximal flow function $f$.
\end{proof}
\textbf{Implementation remark:} In the flow algorithm, while exploring the graph, we explore the neighbours of every vertex in a random order to heuristically spread the association between nodes and partitions.
\subsubsection*{Algorithm}
With this result mind, we can describe the first step of our algorithm. All divisions are supposed to be integer division.
\begin{algorithmic}[1]
\Function{Compute Partition Size}{$\mathbf{N}$, $\mathbf{Z}$, $\mathbf{P}$, $(c_n)_{n\in \mathbf{N}}$, $\rho_\mathbf{N}$, $\rho_\mathbf{Z}$}
\State Build the graph $G=G(s=1)$
\State $ f \leftarrow$ \Call{Maximal flow}{$G$}
\If{$f.\mathrm{total flow} < \rho_\mathbf{N}P$}
\State \Return Error: capacities too small or constraints too strong.
\EndIf
\State $s^- \leftarrow 1$
\State $s^+ \leftarrow 1+\frac{1}{\rho_\mathbf{N}}\sum_{n \in \mathbf{N}} c_n$
\While{$s^-+1 < s^+$}
\State Build the graph $G=G(s=(s^-+s^+)/2)$
\State $ f \leftarrow$ \Call{Maximal flow}{$G$}
\If{$f.\mathrm{total flow} < \rho_\mathbf{N}P$}
\State $s^+ \leftarrow (s^- + s^+)/2$
\Else
\State $s^- \leftarrow (s^- + s^+)/2$
\EndIf
\EndWhile
\State \Return $s^-$
\EndFunction
\end{algorithmic}
\subsubsection*{Complexity}
To compute the maximal flow, we use Dinic's algorithm. Its complexity on general graphs is $O(\#V^2 \#E)$, but on graphs with edge capacity bounded by a constant, it turns out to be $O(\#E^{3/2})$. The graph $G$ does not fall in this case since the capacities of the arcs incoming to $\mathbf{t}$ are far from bounded. However, the proof of this complexity works readily for graph where we only ask the edges \emph{not} incoming to the sink $\mathbf{t}$ to have their capacities bounded by a constant. One can find the proof of this claim in \cite[Section 2]{even1975network}.
The dichotomy adds a logarithmic factor $\log (C)$ where $C=\sum_{n \in \mathbf{N}} c_n$ is the total capacity of the cluster. The total complexity of this first function is hence
$O(\#E^{3/2}\log C ) = O\big((PN)^{3/2} \log C\big)$.
\subsubsection*{Metrics}
We can display the discrepancy between the computed $s^*$ and the best size we could hope for a given total capacity, that is $C/\rho_\mathbf{N}$.
\subsection{Computation of a candidate assignment}
Now that we have the optimal partition size $s^*$, to compute a candidate assignment, it would be enough to compute a maximal flow function $f$ on $G(s^*)$. This is what we do if there was no previous assignment $\alpha'$.
If there was some $\alpha'$, we add a step that will heuristically help to obtain a candidate $\alpha$ closer to $\alpha'$. to do so, we fist compute a flow function $\tilde{f}$ that uses only the partition-to-node association appearing in $\alpha'$. Most likely, $\tilde{f}$ will not be a maximal flow of $G(s^*)$. In Dinic's algorithm, we can start from a non maximal flow function and then discover improving paths. This is what we do in starting from $\tilde{f}$. The hope\footnote{This is only a hope, because one can find examples where the construction of $f$ from $\tilde{f}$ produces an assignment $\alpha$ that is not as close as possible to $\alpha'$.} is that the final flow function $f$ will tend to keep the associations appearing in $\tilde{f}$.
More formally, we construct the graph $G_{|\alpha'}$ from $G$ by removing all the arcs $(\mathbf{x}_{p,z},\mathbf{n}, 1)$ where $p$ is not associated to $n$ in $\alpha'$. We compute a maximal flow function $\tilde{f}$ in $G_{|\alpha'}$. $\tilde{f}$ is also a valid (most likely non maximal) flow function in $G$. We compute a maximal flow function $f$ on $G$ by starting Dinic's algorithm on $\tilde{f}$.
\subsubsection*{Algorithm}
\begin{algorithmic}[1]
\Function{Compute Candidate Assignment}{$G$, $\alpha'$}
\State Build the graph $G_{|\alpha'}$
\State $ \tilde{f} \leftarrow$ \Call{Maximal flow}{$G_{|\alpha'}$}
\State $ f \leftarrow$ \Call{Maximal flow from flow}{$G$, $\tilde{f}$}
\State \Return $f$
\EndFunction
\end{algorithmic}
\textbf{Remark:} The function ``Maximal flow'' can be just seen as the function ``Maximal flow from flow'' called with the zero flow function as starting flow.
\subsubsection*{Complexity}
From the consideration of the last section, we have the complexity of the Dinic's algorithm $O(\#E^{3/2}) = O((PN)^{3/2})$.
\subsubsection*{Metrics}
We can display the flow value of $\tilde{f}$, which is an upper bound of the distance between $\alpha$ and $\alpha'$. It might be more a Debug level display than Info.
\subsection{Minimization of the transfer load}
Now that we have a candidate flow function $f$, we want to modify it to make its associated assignment as close as possible to $\alpha'$. Denote by $f'$ the maximal flow associated to $\alpha'$, and let $d(f, f')$ be distance between the associated assignments\footnote{It is the number of arcs of type $(\mathbf{x}_{p,z},\mathbf{n})$ saturated in one flow and not in the other.}.
We want to build a sequence $f=f_0, f_1, f_2 \dots$ of maximal flows such that $d(f_i, \alpha')$ decreases as $i$ increases. The distance being a non-negative integer, this sequence of flow functions must be finite. We now explain how to find some improving $f_{i+1}$ from $f_i$.
For any maximal flow $f$ in $G$, we define the oriented weighted graph $G_f=(V, E_f)$ as follows. The vertices of $G_f$ are the same as the vertices of $G$. $E_f$ contains the arc $(v_1,v_2, w)$ between vertices $v_1,v_2\in V$ with weight $w$ if and only if the arc $(v_1,v_2)$ is not saturated in $f$ (i.e. $c(v_1,v_2)-f(v_1,v_2) \ge 1$, we also consider reversed arcs). The weight $w$ is:
\begin{itemize}
\item $-1$ if $(v_1,v_2)$ is of type $(\mathbf{x}_{p,z},\mathbf{n})$ or $(\mathbf{x}_{p,z},\mathbf{n})$ and is saturated in only one of the two flows $f,f'$;
\item $+1$ if $(v_1,v_2)$ is of type $(\mathbf{x}_{p,z},\mathbf{n})$ or $(\mathbf{x}_{p,z},\mathbf{n})$ and is saturated in either both or none of the two flows $f,f'$;
\item $0$ otherwise.
\end{itemize}
If $\gamma$ is a simple cycle of arcs in $G_f$, we define its weight $w(\gamma)$ as the sum of the weights of its arcs. We can add $+1$ to the value of $f$ on the arcs of $\gamma$, and by construction of $G_f$ and the fact that $\gamma$ is a cycle, the function that we get is still a valid flow function on $G$, it is maximal as it has the same flow value as $f$. We denote this new function $f+\gamma$.
\begin{proposition}
Given a maximal flow $f$ and a simple cycle $\gamma$ in $G_f$, we have $d(f+\gamma, f') - d(f,f') = w(\gamma)$.
\end{proposition}
\begin{proof}
Let $X$ be the set of arcs of type $(\mathbf{x}_{p,z},\mathbf{n})$. Then we can express $d(f,f')$ as
\begin{align*}
d(f,f') & = \#\{e\in X ~|~ f(e)\neq f'(e)\}
= \sum_{e\in X} 1_{f(e)\neq f'(e)} \\
& = \frac{1}{2}\big( \#X + \sum_{e\in X} 1_{f(e)\neq f'(e)} - 1_{f(e)= f'(e)} \big).
\end{align*}
We can express the cycle weight as
\begin{align*}
w(\gamma) & = \sum_{e\in X, e\in \gamma} - 1_{f(e)\neq f'(e)} + 1_{f(e)= f'(e)}.
\end{align*}
Remark that since we passed on unit of flow in $\gamma$ to construct $f+\gamma$, we have for any $e\in X$, $f(e)=f'(e)$ if and only if $(f+\gamma)(e) \neq f'(e)$.
Hence
\begin{align*}
w(\gamma) & = \frac{1}{2}(w(\gamma) + w(\gamma)) \\
&= \frac{1}{2} \Big(
\sum_{e\in X, e\in \gamma} - 1_{f(e)\neq f'(e)} + 1_{f(e)= f'(e)} \\
& \qquad +
\sum_{e\in X, e\in \gamma} 1_{(f+\gamma)(e)\neq f'(e)} + 1_{(f+\gamma)(e)= f'(e)}
\Big).
\end{align*}
Plugging this in the previous equation, we find that
$$d(f,f')+w(\gamma) = d(f+\gamma, f').$$
\end{proof}
This result suggests that given some flow $f_i$, we just need to find a negative cycle $\gamma$ in $G_{f_i}$ to construct $f_{i+1}$ as $f_i+\gamma$. The following proposition ensures that this greedy strategy reaches an optimal flow.
\begin{proposition}
For any maximal flow $f$, $G_f$ contains a negative cycle if and only if there exists a maximal flow $f^*$ in $G$ such that $d(f^*, f') < d(f, f')$.
\end{proposition}
\begin{proof}
Suppose that there is such flow $f^*$. Define the oriented multigraph $M_{f,f^*}=(V,E_M)$ with the same vertex set $V$ as in $G$, and for every $v_1,v_2 \in V$, $E_M$ contains $(f^*(v_1,v_2) - f(v_1,v_2))_+$ copies of the arc $(v_1,v_2)$. For every vertex $v$, its total degree (meaning its outer degree minus its inner degree) is equal to
\begin{align*}
\deg v & = \sum_{u\in V} (f^*(v,u) - f(v,u))_+ - \sum_{u\in V} (f^*(u,v) - f(u,v))_+ \\
& = \sum_{u\in V} f^*(v,u) - f(v,u) = \sum_{u\in V} f^*(v,u) - \sum_{u\in V} f(v,u).
\end{align*}
The last two sums are zero for any inner vertex since $f,f^*$ are flows, and they are equal on the source and sink since the two flows are both maximal and have hence the same value. Thus, $\deg v = 0$ for every vertex $v$.
This implies that the multigraph $M_{f,f^*}$ is the union of disjoint simple cycles. $f$ can be transformed into $f^*$ by pushing a mass 1 along all these cycles in any order. Since $d(f^*, f')<d(f,f')$, there must exists one of these simple cycles $\gamma$ with $d(f+\gamma, f') < d(f, f')$. Finally, since we can push a mass in $f$ along $\gamma$, it must appear in $G_f$. Hence $\gamma$ is a cycle of $G_f$ with negative weight.
\end{proof}
In the next section we describe the corresponding algorithm. Instead of discovering only one cycle, we are allowed to discover a set $\Gamma$ of disjoint negative cycles.
\subsubsection*{Algorithm}
\begin{algorithmic}[1]
\Function{Minimize transfer load}{$G$, $f$, $\alpha'$}
\State Build the graph $G_f$
\State $\Gamma \leftarrow$ \Call{Detect Negative Cycles}{$G_f$}
\While{$\Gamma \neq \emptyset$}
\ForAll{$\gamma \in \Gamma$}
\State $f \leftarrow f+\gamma$
\EndFor
\State Update $G_f$
\State $\Gamma \leftarrow$ \Call{Detect Negative Cycles}{$G_f$}
\EndWhile
\State \Return $f$
\EndFunction
\end{algorithmic}
\subsubsection*{Complexity}
The distance $d(f,f')$ is bounded by the maximal number of differences in the associated assignment. If these assignment are totally disjoint, this distance is $2\rho_N P$. At every iteration of the While loop, the distance decreases, so there is at most $O(\rho_N P) = O(P)$ iterations.
The detection of negative cycle is done with the Bellman-Ford algorithm, whose complexity should normally be $O(\#E\#V)$. In our case, it amounts to $O(P^2ZN)$. Multiplied by the complexity of the outer loop, it amounts to $O(P^3ZN)$ which is a lot when the number of partitions and nodes starts to be large. To avoid that, we adapt the Bellman-Ford algorithm.
The Bellman-Ford algorithm runs $\#V$ iterations of an outer loop, and an inner loop over $E$. The idea is to compute the shortest paths from a source vertex $v$ to all other vertices. After $k$ iterations of the outer loop, the algorithm has computed all shortest path of length at most $k$. All simple paths have length at most $\#V-1$, so if there is an update in the last iteration of the loop, it means that there is a negative cycle in the graph. The observation that will enable us to improve the complexity is the following:
\begin{proposition}
In the graph $G_f$ (and $G$), all simple paths have a length at most $4N$.
\end{proposition}
\begin{proof}
Since $f$ is a maximal flow, there is no outgoing edge from $\mathbf{s}$ in $G_f$. One can thus check than any simple path of length 4 must contain at least two node of type $\mathbf{n}$. Hence on a path, at most 4 arcs separate two successive nodes of type $\mathbf{n}$.
\end{proof}
Thus, in the absence of negative cycles, shortest paths in $G_f$ have length at most $4N$. So we can do only $4N+1$ iterations of the outer loop in Bellman-Ford algorithm. This makes the complexity of the detection of one set of cycle to be $O(N\#E) = O(N^2 P)$.
With this improvement, the complexity of the whole algorithm is, in the worst case, $O(N^2P^2)$. However, since we detect several cycles at once and we start with a flow that might be close to the previous one, the number of iterations of the outer loop might be smaller in practice.
\subsubsection*{Metrics}
We can display the node and zone utilization ratio, by dividing the flow passing through them divided by their outgoing capacity. In particular, we can pinpoint saturated nodes and zones (i.e. used at their full potential).
We can display the distance to the previous assignment, and the number of partition transfers.
\section{Properties of an optimal 3-strict assignment}
\subsection{Optimal assignment}
\label{sec:opt_assign}
For every zone $z\in Z$, define the zone capacity $c_z = \sum_{v, z_v=z} c_v$ and define $C = \sum_v c_v = \sum_z c_z$.
One can check that the best we could be doing to maximize $s^*$ would be to use the nodes proportionally to their capacity. This would yield $s^*=C/(3N)$. This is not possible because of (i) redundancy constraints and (ii) integer rounding but it gives and upper bound.
\subsubsection*{Optimal utilization}
We call an \emph{utilization} a collection of non-negative integers $(n_v)_{v\in V}$ such that $\sum_v n_v = 3N$ and for every zone $z$, $\sum_{v\in z} n_v \le N$. We call such utilization \emph{optimal} if it maximizes $s^*$.
We start by computing a node sub-utilization $(\hat{n}_v)_{v\in V}$ such that for every zone $z$, $\sum_{v\in z} \hat{n}_v \le N$ and we show that there is an optimal utilization respecting the constraints and such that $\hat{n}_v \le n_v$ for every node.
Assume that there is a zone $z_0$ such that $c_{z_0}/C \ge 1/3$. Then for any $v\in z_0$, we define
$$\hat{n}_v = \left\lfloor\frac{c_v}{c_{z_0}}N\right\rfloor.$$
This choice ensures for any such $v$ that
$$
\frac{c_v}{\hat{n}_v} \ge \frac{c_{z_0}}{N} \ge \frac{C}{3N}
$$
which is the universal upper bound on $s^*$. Hence any optimal utilization $(n_v)$ can be modified to another optimal utilization such that $n_v\ge \hat{n}_v$
Because $z_0$ cannot store more than $N$ partition occurences, in any assignment, at least $2N$ partitions must be assignated to the zones $Z\setminus\{z_0\}$. Let $C_0 = C-c_{z_0}$. Suppose that there exists a zone $z_1\neq z_0$ such that $c_{z_1}/C_0 \ge 1/2$. Then, with the same argument as for $z_0$, we can define
$$\hat{n}_v = \left\lfloor\frac{c_v}{c_{z_1}}N\right\rfloor$$
for every $v\in z_1$.
Now we can assign the remaining partitions. Let $(\hat{N}, \hat{C})$ to be
\begin{itemize}
\item $(3N,C)$ if we did not find any $z_0$;
\item $(2N,C-c_{z_0})$ if there was a $z_0$ but no $z_1$;
\item $(N,C-c_{z_0}-c_{z_1})$ if there was a $z_0$ and a $z_1$.
\end{itemize}
Then at least $\hat{N}$ partitions must be spread among the remaining zones. Hence $s^*$ is upper bounded by $\hat{C}/\hat{N}$ and without loss of generality, we can define, for every node that is not in $z_0$ nor $z_1$,
$$\hat{n}_v = \left\lfloor\frac{c_v}{\hat{C}}\hat{N}\right\rfloor.$$
We constructed a sub-utilization $\hat{n}_v$. Now notice that $3N-\sum_v \hat{n}_v \le \# V$ where $\# V$ denotes the number of nodes. We can iteratively pick a node $v^*$ such that
\begin{itemize}
\item $\sum_{v\in z_{v^*}} \hat{n}_v < N$ where $z_{v^*}$ is the zone of $v^*$;
\item $v^*$ maximizes the quantity $c_v/(\hat{n}_v+1)$ among the vertices satisfying the first condition (i.e. not in a saturated zone).
\end{itemize}
We iterate these instructions until $\sum_v \hat{n}_v= 3N$, and at this stage we define $(n_v) = (\hat{n}_v)$. It is easy to prove by induction that at every step, there is an optimal utilization that is pointwise larger than $\hat{n}_v$, and in particular, that $(n_v)$ is optimal.
\subsubsection*{Existence of an optimal assignment}
As for now, the \emph{optimal utilization} that we obtained is just a vector of numbers and it is not clear that it can be realized as the utilization of some concrete assignment. Here is a way to get a concrete assignment.
Define $3N$ tokens $t_1,\ldots, t_{3N}\in V$ as follows:
\begin{itemize}
\item Enumerate the zones $z$ of $Z$ in any order;
\item enumerate the nodes $v$ of $z$ in any order;
\item repeat $n_v$ times the token $v$.
\end{itemize}
Then for $1\le i \le N$, define the triplet $T_i$ to be
$(t_i, t_{i+N}, t_{i+2N})$. Since the same nodes of a zone appear contiguously, the three nodes of a triplet must belong to three distinct zones.
However simple, this solution to go from an utilization to an assignment has the drawback of not spreading the triplets: a node will tend to be associated to the same two other nodes for many partitions. Hence, during data transfer, it will tend to use only two link, instead of spreading the bandwith use over many other links to other nodes. To achieve this goal, we will reframe the search of an assignment as a flow problem. and in the flow algorithm, we will introduce randomness in the order of exploration. This will be sufficient to obtain a good dispersion of the triplets.
\begin{figure}
\centering
\includegraphics[width=0.9\linewidth]{figures/naive}
\caption{On the left, the creation of a concrete assignment with the naive approach of repeating tokens. On the right, the zones containing the nodes.}
\end{figure}
\subsubsection*{Assignment as a maximum flow problem}
We describe the flow problem via its graph $(X,E)$ where $X$ is a set of vertices, and $E$ are directed weighted edges between the vertices. For every zone $z$, define $n_z=\sum_{v\in z} n_v$.
The set of vertices $X$ contains the source $\mathbf{s}$ and the sink $\mathbf{t}$; a vertex $\mathbf{x}_z$ for every zone $z\in Z$, and a vertex $\mathbf{y}_i$ for every partition index $1\le i\le N$.
The set of edges $E$ contains
\begin{itemize}
\item the edge $(\mathbf{s}, \mathbf{x}_z, n_z)$ for every zone $z\in Z$;
\item the edge $(\mathbf{x}_z, \mathbf{y}_i, 1)$ for every zone $z\in Z$ and partition $1\le i\le N$;
\item the edge $(\mathbf{y}_i, \mathbf{t}, 3)$ for every partition $1\le i\le N$.
\end{itemize}
\begin{figure}[b]
\centering
\includegraphics[width=0.6\linewidth]{figures/flow}
\caption{Flow problem to compute and optimal assignment.}
\end{figure}
We first show the equivalence between this problem and and the construction of an assignment. Given some optimal assignment $(n_v)$, define the flow $f:E\to \mathbb{N}$ that saturates every edge from $\mathbf{s}$ or to $\mathbf{t}$, takes value $1$ on the edge between $\mathbf{x}_z$ and $\mathbf{y}_i$ if partition $i$ is stored in some node of the zone $z$, and $0$ otherwise. One can easily check that $f$ thus defined is indeed a flow and is maximum.
Reciprocally, by the existence of maximum flows constructed from optimal assignments, any maximum flow must saturate the edges linked to the source or the sink. It can only take value 0 or 1 on the other edge, and every partition vertex is associated to exactly three distinct zone vertices. Every zone is associated to exactly $n_z$ partitions.
A maximum flow can be constructed using, for instance, Dinic's algorithm. This algorithm works by discovering augmenting path to iteratively increase the flow. During the exploration of the graph to find augmenting path, we can shuffle the order of enumeration of the neighbours to spread the associations between zones and partitions.
Once we have such association, we can randomly distribute the $n_z$ edges picked for every zone $z$ to its nodes $v\in z$ such that every such $v$ gets $n_z$ edges. This defines an optimal assignment of partitions to nodes.
\subsection{Minimal transfer}
Assume that there was a previous assignment $(T'_i)_{1\le i\le N}$ corresponding to utilizations $(n'_v)_{v\in V}$. We would like the new computed assignment $(T_i)_{1\le i\le N}$ from some $(n_v)_{v\in V}$ to minimize the number of partitions that need to be transferred. We can imagine two different objectives corresponding to different hypotheses:
\begin{equation}
\tag{H3A}
\label{hyp:A}
\text{\emph{Transfers between different zones cost much more than inside a zone.}}
\end{equation}
\begin{equation}
\tag{H3B}
\label{hyp:B}
\text{\emph{Changing zone is not the largest cost when transferring a partition.}}
\end{equation}
In case $A$, our goal will be to minimize the number of changes of zone in the assignment of partitions to zone. More formally, we will maximize the quantity
$$
Q_Z :=
\sum_{1\le i\le N}
\#\{z\in Z ~|~ z\cap T_i \neq \emptyset, z\cap T'_i \neq \emptyset \}
.$$
In case $B$, our goal will be to minimize the number of changes of nodes in the assignment of partitions to nodes. We will maximize the quantity
$$
Q_V :=
\sum_{1\le i\le N} \#(T_i \cap T'_i).
$$
It is tempting to hope that there is a way to maximize both quantity, that having the least discrepancy in terms of nodes will lead to the least discrepancy in terms of zones. But this is actually wrong! We propose the following counter-example to convince the reader:
We consider eight nodes $a, a', b, c, d, d', e, e'$ belonging to five different zones $\{a,a'\}, \{b\}, \{c\}, \{d,d'\}, \{e, e'\}$. We take three partitions ($N=3$), that are originally assigned with some utilization $(n'_v)_{v\in V}$ as follows:
$$
T'_1=(a,b,c) \qquad
T'_2=(a',b,d) \qquad
T'_3=(b,c,e).
$$
This assignment, with updated utilizations $(n_v)_{v\in V}$ minimizes the number of zone changes:
$$
T_1=(d,b,c) \qquad
T_2=(a,b,d) \qquad
T_3=(b,c,e').
$$
This one, with the same utilization, minimizes the number of node changes:
$$
T_1=(a,b,c) \qquad
T_2=(e',b,d) \qquad
T_3=(b,c,d').
$$
One can check that in this case, it is impossible to minimize both the number of zone and node changes.
Because of the redundancy constraint, we cannot use a greedy algorithm to just replace nodes in the triplets to try to get the new utilization rate: this could lead to blocking situation where there is still a hole to fill in a triplet but no available node satisfies the zone separation constraint. To circumvent this issue, we propose an algorithm based on finding cycles in a graph encoding of the assignment. As in section \ref{sec:opt_assign}, we can explore the neigbours in a random order in the graph algorithms, to spread the triplets distribution.
\subsubsection{Minimizing the zone discrepancy}
First, notice that, given an assignment of partitions to \emph{zones}, it is easy to deduce an assignment to \emph{nodes} that minimizes the number of transfers for this zone assignment: For every zone $z$ and every node $v\in z$, pick in any way a set $P_v$ of partitions that where assigned to $v$ in $T'$, to $z_v$ in $T$, with the cardinality of $P_v$ smaller than $n_v$. Once all these sets are chosen, complement the assignment to reach the right utilization for every node. If $\#P_v > n_v$, it means that all the partitions that could stay in $v$ (i.e. that were already in $v$ and are still assigned to its zone) do stay in $v$. If $\#P_v = n_v$, then $n_v$ partitions stay in $v$, which is the number of partitions that need to be in $v$ in the end. In both cases, we could not hope for better given the partition to zone assignment.
Our goal now is to find a assignment of partitions to zones that minimizes the number of zone transfers. To do so we are going to represent an assignment as a graph.
Let $G_T=(X,E_T)$ be the directed weighted graph with vertices $(\mathbf{x}_i)_{1\le i\le N}$ and $(\mathbf{y}_z)_{z\in Z}$. For any $1\le i\le N$ and $z\in Z$, $E_T$ contains the arc:
\begin{itemize}
\item $(\mathbf{x}_i, \mathbf{y}_z, +1)$, if $z$ appears in $T_i'$ and $T_i$;
\item $(\mathbf{x}_i, \mathbf{y}_z, -1)$, if $z$ appears in $T_i$ but not in $T'_i$;
\item $(\mathbf{y}_z, \mathbf{x}_i, -1)$, if $z$ appears in $T'_i$ but not in $T_i$;
\item $(\mathbf{y}_z, \mathbf{x}_i, +1)$, if $z$ does not appear in $T'_i$ nor in $T_i$.
\end{itemize}
In other words, the orientation of the arc encodes whether partition $i$ is stored in zone $z$ in the assignment $T$ and the weight $\pm 1$ encodes whether this corresponds to what happens in the assignment $T'$.
\begin{figure}[t]
\centering
\begin{minipage}{.40\linewidth}
\centering
\includegraphics[width=.8\linewidth]{figures/mini_zone}
\end{minipage}
\begin{minipage}{.55\linewidth}
\centering
\includegraphics[width=.8\linewidth]{figures/mini_node}
\end{minipage}
\caption{On the left: the graph $G_T$ encoding an assignment to minimize the zone discrepancy. On the right: the graph $G_T$ encoding an assignment to minimize the node discrepancy.}
\end{figure}
Notice that at every partition, there are three outgoing arcs, and at every zone, there are $n_z$ incoming arcs. Moreover, if $w(e)$ is the weight of an arc $e$, define the weight of $G_T$ by
\begin{align*}
w(G_T) := \sum_{e\in E} w(e) &= \#Z \times N - 4 \sum_{1\le i\le N} \#\{z\in Z ~|~ z\cap T_i = \emptyset, z\cap T'_i \neq \emptyset\} \\
&=\#Z \times N - 4 \sum_{1\le i\le N} 3- \#\{z\in Z ~|~ z\cap T_i \neq \emptyset, z\cap T'_i \neq \emptyset\} \\
&= (\#Z-12)N + 4 Q_Z.
\end{align*}
Hence maximizing $Q_Z$ is equivalent to maximizing $w(G_T)$.
Assume that their exist some assignment $T^*$ with the same utilization $(n_v)_{v\in V}$. Define $G_{T^*}$ similarly and consider the set $E_\mathrm{Diff} = E_T \setminus E_{T^*}$ of arcs that appear only in $G_T$. Since all vertices have the same number of incoming arcs in $G_T$ and $G_{T^*}$, the vertices of the graph $(X, E_\mathrm{Diff})$ must all have the same number number of incoming and outgoing arrows. So $E_\mathrm{Diff}$ can be expressed as a union of disjoint cycles. Moreover, the edges of $E_\mathrm{Diff}$ must appear in $E_{T^*}$ with reversed orientation and opposite weight. Hence, we have
$$
w(G_T) - w(G_{T^*}) = 2 \sum_{e\in E_\mathrm{Diff}} w(e).
$$
Hence, if $T$ is not optimal, there exists some $T^*$ with $w(G_T) < w(G_{T^*})$, and by the considerations above, there must exist a cycle in $E_\mathrm{Diff}$, and hence in $G_T$, with negative weight. If we reverse the edges and weights along this cycle, we obtain some graph. Since we did not change the incoming degree of any vertex, this is the graph encoding of some valid assignment $T^+$ such that $w(G_{T^+}) > w(G_T)$. We can iterate this operation until there is no other assignment $T^*$ with larger weight, that is until we obtain an optimal assignment.
\subsubsection{Minimizing the node discrepancy}
We will follow an approach similar to the one where we minimize the zone discrepancy. Here we will directly obtain a node assignment from a graph encoding.
Let $G_T=(X,E_T)$ be the directed weighted graph with vertices $(\mathbf{x}_i)_{1\le i\le N}$, $(\mathbf{y}_{z,i})_{z\in Z, 1\le i\le N}$ and $(\mathbf{u}_v)_{v\in V}$. For any $1\le i\le N$ and $z\in Z$, $E_T$ contains the arc:
\begin{itemize}
\item $(\mathbf{x}_i, \mathbf{y}_{z,i}, 0)$, if $z$ appears in $T_i$;
\item $(\mathbf{y}_{z,i}, \mathbf{x}_i, 0)$, if $z$ does not appear in $T_i$.
\end{itemize}
For any $1\le i\le N$ and $v\in V$, $E_T$ contains the arc:
\begin{itemize}
\item $(\mathbf{y}_{z_v,i}, \mathbf{u}_v, +1)$, if $v$ appears in $T_i'$ and $T_i$;
\item $(\mathbf{y}_{z_v,i}, \mathbf{u}_v, -1)$, if $v$ appears in $T_i$ but not in $T'_i$;
\item $(\mathbf{u}_v, \mathbf{y}_{z_v,i}, -1)$, if $v$ appears in $T'_i$ but not in $T_i$;
\item $(\mathbf{u}_v, \mathbf{y}_{z_v,i}, +1)$, if $v$ does not appear in $T'_i$ nor in $T_i$.
\end{itemize}
Every vertex $\mathbb{x}_i$ has outgoing degree 3, every vertex $\mathbf{y}_{z,v}$ has outgoing degree 1, and every vertex $\mathbf{u}_v$ has incoming degree $n_v$.
Remark that any graph respecting these degree constraints is the encoding of a valid assignment with utilizations $(n_v)_{v\in V}$, in particular no partition is stored in two nodes of the same zone.
We define $w(G_T)$ similarly:
\begin{align*}
w(G_T) := \sum_{e\in E_T} w(e) &= \#V \times N - 4\sum_{1\le i\le N} 3-\#(T_i\cap T'_i) \\
&= (\#V-12)N + 4Q_V.
\end{align*}
Exactly like in the previous section, the existence of an assignment with larger weight implies the existence of a negatively weighted cycle in $G_T$. Reversing this cycle gives us the encoding of a valid assignment with a larger weight. Iterating this operation yields an optimal assignment.
\subsubsection{Linear combination of both criteria}
In the graph $G_T$ defined in the previous section, instead of having weights $0$ and $\pm 1$, we could be having weights $\pm\alpha$ between $\mathbf{x}$ and $\mathbf{y}$ vertices, and weights $\pm\beta$ between $\mathbf{y}$ and $\mathbf{u}$ vertices, for some $\alpha,\beta>0$ (we have positive weight if the assignment corresponds to $T'$ and negative otherwise). Then
\begin{align*}
w(G_T) &= \sum_{e\in E_T} w(e) =
\alpha \big( (\#Z-12)N + 4 Q_Z\big) +
\beta \big( (\#V-12)N + 4 Q_V\big) \\
&= \mathrm{const}+ 4(\alpha Q_Z + \beta Q_V).
\end{align*}
So maximizing the weight of such graph encoding would be equivalent to maximizing a linear combination of $Q_Z$ and $Q_V$.
\subsection{Algorithm}
We give a high level description of the algorithm to compute an optimal 3-strict assignment. The operations appearing at lines 1,2,4 are respectively described by Algorithms \ref{alg:util},\ref{alg:opt} and \ref{alg:mini}.
\begin{algorithm}[H]
\caption{Optimal 3-strict assignment}
\label{alg:total}
\begin{algorithmic}[1]
\Function{Optimal 3-strict assignment}{$N$, $(c_v)_{v\in V}$, $T'$}
\State $(n_v)_{v\in V} \leftarrow$ \Call{Compute optimal utilization}{$N$, $(c_v)_{v\in V}$}
\State $(T_i)_{1\le i\le N} \leftarrow$ \Call{Compute candidate assignment}{$N$, $(n_v)_{v\in V}$}
\If {there was a previous assignment $T'$}
\State $T \leftarrow$ \Call{Minimization of transfers}{$(T_i)_{1\le i\le N}$, $(T'_i)_{1\le i\le N}$}
\EndIf
\State \Return $T$.
\EndFunction
\end{algorithmic}
\end{algorithm}
We give some considerations of worst case complexity for these algorithms. In the following, we assume $N>\#V>\#Z$. The complexity of Algorithm \ref{alg:total} is $O(N^3\# Z)$ if we assume \eqref{hyp:A} and $O(N^3 \#Z \#V)$ if we assume \eqref{hyp:B}.
Algorithm \ref{alg:util} can be implemented with complexity $O(\#V^2)$. The complexity of the function call at line \ref{lin:subutil} is $O(\#V)$. The difference between the sum of the subutilizations and $3N$ is at most the sum of the rounding errors when computing the $\hat{n}_v$. Hence it is bounded by $\#V$ and the loop at line \ref{lin:loopsub} is iterated at most $\#V$ times. Finding the minimizing $v$ at line \ref{lin:findmin} takes $O(\#V)$ operations (naively, we could also use a heap).
Algorithm \ref{alg:opt} can be implemented with complexity $O(N^3\times \#Z)$. The flow graph has $O(N+\#Z)$ vertices and $O(N\times \#Z)$ edges. Dinic's algorithm has complexity $O(\#\mathrm{Vertices}^2\#\mathrm{Edges})$ hence in our case it is $O(N^3\times \#Z)$.
Algorithm \ref{alg:mini} can be implented with complexity $O(N^3\# Z)$ under \eqref{hyp:A} and $O(N^3 \#Z \#V)$ under \eqref{hyp:B}.
The graph $G_T$ has $O(N)$ vertices and $O(N\times \#Z)$ edges under assumption \eqref{hyp:A} and respectively $O(N\times \#Z)$ vertices and $O(N\times \#V)$ edges under assumption \eqref{hyp:B}. The loop at line \ref{lin:repeat} is iterated at most $N$ times since the distance between $T$ and $T'$ decreases at every iteration. Bellman-Ford algorithm has complexity $O(\#\mathrm{Vertices}\#\mathrm{Edges})$, which in our case amounts to $O(N^2\# Z)$ under \eqref{hyp:A} and $O(N^2 \#Z \#V)$ under \eqref{hyp:B}.
\begin{algorithm}
\caption{Computation of the optimal utilization}
\label{alg:util}
\begin{algorithmic}[1]
\Function{Compute optimal utilization}{$N$, $(c_v)_{v\in V}$}
\State $(\hat{n}_v)_{v\in V} \leftarrow $ \Call{Compute subutilization}{$N$, $(c_v)_{v\in V}$} \label{lin:subutil}
\While{$\sum_{v\in V} \hat{n}_v < 3N$} \label{lin:loopsub}
\State Pick $v\in V$ minimizing $\frac{c_v}{\hat{n}_v+1}$ and such that
$\sum_{v'\in z_v} \hat{n}_{v'} < N$ \label{lin:findmin}
\State $\hat{n}_v \leftarrow \hat{n}_v+1$
\EndWhile
\State \Return $(\hat{n}_v)_{v\in V}$
\EndFunction
\State
\Function{Compute subutilization}{$N$, $(c_v)_{v\in V}$}
\State $R \leftarrow 3$
\For{$v\in V$}
\State $\hat{n}_v \leftarrow \mathrm{unset}$
\EndFor
\For{$z\in Z$}
\State $c_z \leftarrow \sum_{v\in z} c_v$
\EndFor
\State $C \leftarrow \sum_{z\in Z} c_z$
\While{$\exists z \in Z$ such that $R\times c_{z} > C$}
\For{$v\in z$}
\State $\hat{n}_v \leftarrow \left\lfloor \frac{c_v}{c_z} N \right\rfloor$
\EndFor
\State $C \leftarrow C-c_z$
\State $R\leftarrow R-1$
\EndWhile
\For{$v\in V$}
\If{$\hat{n}_v = \mathrm{unset}$}
\State $\hat{n}_v \leftarrow \left\lfloor \frac{Rc_v}{C} N \right\rfloor$
\EndIf
\EndFor
\State \Return $(\hat{n}_v)_{v\in V}$
\EndFunction
\end{algorithmic}
\end{algorithm}
\begin{algorithm}
\caption{Computation of a candidate assignment}
\label{alg:opt}
\begin{algorithmic}[1]
\Function{Compute candidate assignment}{$N$, $(n_v)_{v\in V}$}
\State Compute the flow graph $G$
\State Compute the maximal flow $f$ using Dinic's algorithm with randomized neighbours enumeration
\State Construct the assignment $(T_i)_{1\le i\le N}$ from $f$
\State \Return $(T_i)_{1\le i\le N}$
\EndFunction
\end{algorithmic}
\end{algorithm}
\begin{algorithm}
\caption{Minimization of the number of transfers}
\label{alg:mini}
\begin{algorithmic}[1]
\Function{Minimization of transfers}{$(T_i)_{1\le i\le N}$, $(T'_i)_{1\le i\le N}$}
\State Construct the graph encoding $G_T$
\Repeat \label{lin:repeat}
\State Find a negative cycle $\gamma$ using Bellman-Ford algorithm on $G_T$
\State Reverse the orientations and weights of edges in $\gamma$
\Until{no negative cycle is found}
\State Update $(T_i)_{1\le i\le N}$ from $G_T$
\State \Return $(T_i)_{1\le i\le N}$
\EndFunction
\end{algorithmic}
\end{algorithm}
\newpage
\section{Computation of a 3-non-strict assignment}
\subsection{Choices of optimality}
In this mode, we primarily want to store every partition on three nodes, and only secondarily try to spread the nodes among different zone. So we make the choice of not taking the zone repartition in the criterion of optimality.
We try to maximize $s^*$ defined in \eqref{eq:optimal}. So we can compute the optimal utilizations $(n_v)_{v\in V}$ with the only constraint that $n_v \le N$ for every node $v$. As in the previous section, we start with a sub-utilization proportional to $c_v$ (and capped at $N$), and we iteratively increase the $\hat{n}_v$ that is less than $N$ and maximizes the quantity $c_v/(\hat{n}_v+1)$, until the total sum is $3N$.
\subsection{Computation of a candidate assignment}
To compute a candidate assignment (that does not optimize zone spreading nor distance to a previous assignment yet), we can use the folowing flow problem.
Define the oriented weighted graph $(X,E)$. The set of vertices $X$ contains the source $\mathbf{s}$, the sink $\mathbf{t}$, vertices
$\mathbf{x}_p, \mathbf{u}^+_p, \mathbf{u}^-_p$ for every partition $p$, vertices $\mathbf{y}_{p,z}$ for every partition $p$ and zone $z$, and vertices $\mathbf{z}_v$ for every node $v$.
The set of edges is composed of the following arcs:
\begin{itemize}
\item ($\mathbf{s}$,$\mathbf{x}_p$, 3) for every partition $p$;
\item ($\mathbf{x}_p$,$\mathbf{u}^+_p$, 3) for every partition $p$;
\item ($\mathbf{x}_p$,$\mathbf{u}^-_p$, 2) for every partition $p$;
\item ($\mathbf{u}^+_p$,$\mathbf{y}_{p,z}$, 1) for every partition $p$ and zone $z$;
\item ($\mathbf{u}^-_p$,$\mathbf{y}_{p,z}$, 2) for every partition $p$ and zone $z$;
\item ($\mathbf{y}_{p,z}$,$\mathbf{z}_v$, 1) for every partition $p$, zone $z$ and node $v\in z$;
\item ($\mathbf{z}_v$, $\mathbf{t}$, $n_v$) for every node $v$;
\end{itemize}
One can check that any maximal flow in this graph corresponds to an assignment of partitions to nodes. In such a flow, all the arcs from $\mathbf{s}$ and to $\mathbf{t}$ are saturated. The arc from $\mathbf{y}_{p,z}$ to $\mathbf{z}_v$ is saturated if and only if $p$ is associated to~$v$.
Finally the flow from $\mathbf{x}_p$ to $\mathbf{y}_{p,z}$ can go either through $\mathbf{u}^+_p$ or $\mathbf{u}^-_p$.
\subsection{Maximal spread and minimal transfers}
Notice that if the arc $\mathbf{u}_p^+\mathbf{y}_{p,z}$ is not saturated but there is some flow in $\mathbf{u}_p^-\mathbf{y}_{p,z}$, then it is possible to transfer a unit of flow from the path $\mathbf{x}_p\mathbf{u}_p^-\mathbf{y}_{p,z}$ to the path $\mathbf{x}_p\mathbf{u}_p^+\mathbf{y}_{p,z}$. So we can always find an equivalent maximal flow $f^*$ that uses the path through $\mathbf{u}_p^-$ only if the path through $\mathbf{u}_p^+$ is saturated.
We will use this fact to consider the amount of flow going through the vertices $\mathbf{u}^+$ as a measure of how well the partitions are spread over nodes belonging to different zones. If the partition $p$ is associated to 3 different zones, then a flow of 3 will cross $\mathbf{u}_p^+$ in $f^*$ (i.e. a flow of 0 will cross $\mathbf{u}_p^+$). If $p$ is associated to two zones, a flow of $2$ will cross $\mathbf{u}_p^+$. If $p$ is associated to a single zone, a flow of $1$ will cross $\mathbf{u}_p^+$.
Let $N_1, N_2, N_3$ be the number of partitions associated to respectively 1,2 and 3 distinct zones. We will optimize a linear combination of these variables using the discovery of positively weighted circuits in a graph.
At the same step, we will also optimize the distance to a previous assignment $T'$. Let $\alpha> \beta> \gamma \ge 0$ be three parameters.
Given the flow $f$, let $G_f=(X',E_f)$ be the multi-graph where $X' = X\setminus\{\mathbf{s},\mathbf{t}\}$. The set $E_f$ is composed of the arcs:
\begin{itemize}
\item As many arcs from $(\mathbf{x}_p, \mathbf{u}^+_p,\alpha), (\mathbf{x}_p, \mathbf{u}^+_p,\beta), (\mathbf{x}_p, \mathbf{u}^+_p,\gamma)$ (selected in this order) as there is flow crossing $\mathbf{u}^+_p$ in $f$;
\item As many arcs from $(\mathbf{u}^+_p, \mathbf{x}_p,-\gamma), (\mathbf{u}^+_p, \mathbf{x}_p,-\beta), (\mathbf{u}^+_p, \mathbf{x}_p,-\alpha)$ (selected in this order) as there is flow crossing $\mathbf{u}^-_p$ in $f$;
\item As many copies of $(\mathbf{x}_p, \mathbf{u}^-_p,0)$ as there is flow through $\mathbf{u}^-_p$;
\item As many copies of $(\mathbf{u}^-_p,\mathbf{x}_p,0)$ so that the number of arcs between these two vertices is 2;
\item $(\mathbf{u}^+_p,\mathbf{y}_{p,z}, 0)$ if the flow between these vertices is 1, and the opposite arc otherwise;
\item as many copies of $(\mathbf{u}^-_p,\mathbf{y}_{p,z}, 0)$ as the flow between these vertices, and as many copies of the opposite arc as 2~$-$~the flow;
\item $(\mathbf{y}_{p,z},\mathbf{z}_v, \pm1)$ if it is saturated in $f$, with $+1$ if $v\in T'_p$ and $-1$ otherwise;
\item $(\mathbf{z}_v,\mathbf{y}_{p,z}, \pm1)$ if it is not saturated in $f$, with $+1$ if $v\notin T'_p$ and $-1$ otherwise.
\end{itemize}
To summarize, arcs are oriented left to right if they correspond to a presence of flow in $f$, and right to left if they correspond to an absence of flow. They are positively weighted if we want them to stay at their current state, and negatively if we want them to switch. Let us compute the weight of such graph.
\begin{multline*}
w(G_f) = \sum_{e\in E_f} w(e_f) \\
=
(\alpha - \beta -\gamma) N_1 + (\alpha +\beta - \gamma) N_2 + (\alpha+\beta+\gamma) N_3
\\ +
\#V\times N - 4 \sum_p 3-\#(T_p\cap T'_p) \\
=(\#V-12+\alpha-\beta-\gamma)\times N + 4Q_V + 2\beta N_2 + 2(\beta+\gamma) N_3 \\
\end{multline*}
As for the mode 3-strict, one can check that the difference of two such graphs corresponding to the same $(n_v)$ is always eulerian. Hence we can navigate in this class with the same greedy algorithm that discovers positive cycles and flips them.
The function that we optimize is
$$
2Q_V + \beta N_2 + (\beta+\gamma) N_3.
$$
The choice of parameters $\beta$ and $\gamma$ should be lead by the following question: For $\beta$, where to put the tradeoff between zone dispersion and distance to the previous configuration? For $\gamma$, do we prefer to have more partitions spread between 2 zones, or have less between at least 2 zones but more between 3 zones.
The quantity $Q_V$ varies between $0$ and $3N$, it should be of order $N$. The quantity $N_2+N_3$ should also be of order $N$ (it is exactly $N$ in the strict mode). So the two terms of the function are comparable.
\bibliography{optimal_layout}
\bibliographystyle{ieeetr}
\end{document}

BIN
doc/sticker/Garage.pdf Normal file

Binary file not shown.

BIN
doc/sticker/Garage.png Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 36 KiB

Binary file not shown.

1
doc/talks/.gitignore vendored Normal file
View file

@ -0,0 +1 @@
.direnv/

17
doc/talks/2023-09-20-ocp/.gitignore vendored Normal file
View file

@ -0,0 +1,17 @@
*
!*.txt
!*.md
!assets
!.gitignore
!*.svg
!*.png
!*.jpg
!*.tex
!Makefile
!.gitignore
!assets/*.drawio.pdf
!talk.pdf

View file

@ -0,0 +1,34 @@
ASSETS=assets/consistent_hashing_1.pdf \
assets/consistent_hashing_2.pdf \
assets/consistent_hashing_3.pdf \
assets/consistent_hashing_4.pdf \
assets/garage_tables.pdf \
assets/consensus.pdf_tex \
assets/lattice1.pdf_tex \
assets/lattice2.pdf_tex \
assets/lattice3.pdf_tex \
assets/lattice4.pdf_tex \
assets/lattice5.pdf_tex \
assets/lattice6.pdf_tex \
assets/lattice7.pdf_tex \
assets/lattice8.pdf_tex \
assets/latticeB_1.pdf_tex \
assets/latticeB_2.pdf_tex \
assets/latticeB_3.pdf_tex \
assets/latticeB_4.pdf_tex \
assets/latticeB_5.pdf_tex \
assets/latticeB_6.pdf_tex \
assets/latticeB_7.pdf_tex \
assets/latticeB_8.pdf_tex \
assets/latticeB_9.pdf_tex \
assets/latticeB_10.pdf_tex \
assets/deuxfleurs.pdf
talk.pdf: talk.tex $(ASSETS)
pdflatex talk.tex
assets/%.pdf: assets/%.svg
inkscape -D -z --file=$^ --export-pdf=$@
assets/%.pdf_tex: assets/%.svg
inkscape -D -z --file=$^ --export-pdf=$@ --export-latex

View file

@ -0,0 +1,39 @@
### (fr) Garage, un système de stockage de données géo-distribué léger et robuste
Garage est un système de stockage de données léger, géo-distribué, qui
implémente le protocole de stockage S3 de Amazon. Garage est destiné
principalement à l'auto-hébergement sur du matériel courant d'occasion. À ce
titre, il doit tolérer un grand nombre de pannes: coupures de courant, coupures
de connexion Internet, pannes de machines, ... Il doit également être facile à
déployer et à maintenir, afin de pouvoir être facilement utilisé par des
amateurs ou des petites organisations.
Cette présentation vous proposera un aperçu de Garage et du choix technique
principal qui rend un système comme Garage possible: le refus d'utiliser des
algorithmes de consensus, remplacés avantageusement par des méthodes à
cohérence faible. Notre modèle est fortement inspiré de la base de donnée
Dynamo (DeCandia et al, 2007), et fait usage des types de données CRDT (Shapiro
et al, 2011). Nous exploreront comment ces méthodes s'appliquent à la
construction de l'abstraction "stockage objet" dans un système distribué, et
quelles autres abstractions peuvent ou ne peuvent pas être construites dans ce
modèle.
### (en) Garage, a lightweight and robust geo-distributed data storage system
Garage is a lightweight geo-distributed data store that implements the Amazon
S3 object storage protocol. Garage is meant primarily for self-hosting at home
on second-hand commodity hardware, meaning it has to tolerate a wide variety of
failure scenarios such as power cuts, Internet disconnections and machine
crashes. It also has to be easy to deploy and maintain, so that hobbyists and
small organizations can use it without trouble.
This talk will present Garage and the key technical choice that made Garage
possible: refusing to use consensus algorithms and using instead weak
consistency methods, with a model that is loosely based on that of the Dynamo
database (DeCandia et al, 2007) and that makes heavy use of conflict-free
replicated data types (Shapiro et al, 2011). We will explore how these methods
are suited to building the "object store" abstraction in a distributed system,
and what other abstractions are possible or impossible to build in this model.

Binary file not shown.

After

Width:  |  Height:  |  Size: 32 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 115 KiB

File diff suppressed because one or more lines are too long

After

Width:  |  Height:  |  Size: 184 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 26 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 27 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 8.9 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 18 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 4.8 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 263 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 82 KiB

View file

@ -0,0 +1,137 @@
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!-- Created with Inkscape (http://www.inkscape.org/) -->
<svg
width="800"
height="300"
viewBox="0 0 211.66666 79.374999"
version="1.1"
id="svg5"
inkscape:version="1.2.2 (b0a8486541, 2022-12-01)"
sodipodi:docname="consensus.svg"
xmlns:inkscape="http://www.inkscape.org/namespaces/inkscape"
xmlns:sodipodi="http://sodipodi.sourceforge.net/DTD/sodipodi-0.dtd"
xmlns="http://www.w3.org/2000/svg"
xmlns:svg="http://www.w3.org/2000/svg">
<sodipodi:namedview
id="namedview7"
pagecolor="#ffffff"
bordercolor="#666666"
borderopacity="1.0"
inkscape:showpageshadow="2"
inkscape:pageopacity="0.0"
inkscape:pagecheckerboard="0"
inkscape:deskcolor="#d1d1d1"
inkscape:document-units="mm"
showgrid="false"
inkscape:zoom="1.4734708"
inkscape:cx="310.49139"
inkscape:cy="179.1688"
inkscape:window-width="1920"
inkscape:window-height="999"
inkscape:window-x="0"
inkscape:window-y="0"
inkscape:window-maximized="1"
inkscape:current-layer="layer1" />
<defs
id="defs2">
<marker
style="overflow:visible"
id="Arrow2"
refX="0"
refY="0"
orient="auto-start-reverse"
inkscape:stockid="Arrow2"
markerWidth="7.6999998"
markerHeight="5.5999999"
viewBox="0 0 7.7 5.6"
inkscape:isstock="true"
inkscape:collect="always"
preserveAspectRatio="xMidYMid">
<path
transform="scale(0.7)"
d="M -2,-4 9,0 -2,4 c 2,-2.33 2,-5.66 0,-8 z"
style="fill:context-stroke;fill-rule:evenodd;stroke:none"
id="arrow2L" />
</marker>
</defs>
<g
inkscape:label="Layer 1"
inkscape:groupmode="layer"
id="layer1">
<g
id="g1218"
transform="translate(-8.9161476,-12.502301)">
<circle
style="fill:#ffffff;stroke:#000000;stroke-width:1;stroke-dasharray:none;stroke-opacity:1"
id="path111"
cx="38.904896"
cy="37.936272"
r="13.474442" />
<text
xml:space="preserve"
style="font-size:8.46667px;line-height:1.25;font-family:sans-serif;text-align:center;text-anchor:middle;stroke-width:0.264583;fill:#000000"
x="38.879501"
y="40.908073"
id="text1105"><tspan
sodipodi:role="line"
id="tspan1103"
style="stroke-width:0.264583;fill:#000000"
x="38.879501"
y="40.908073">$\bot$</tspan></text>
</g>
<g
id="g1218-3"
transform="translate(127.41938,-12.502301)">
<circle
style="fill:#ffffff;stroke:#000000;stroke-width:1;stroke-dasharray:none;stroke-opacity:1"
id="path111-5"
cx="38.904896"
cy="37.936272"
r="13.474442" />
<text
xml:space="preserve"
style="font-size:8.46667px;line-height:1.25;font-family:sans-serif;text-align:center;text-anchor:middle;stroke-width:0.264583;fill:#000000"
x="38.879501"
y="40.908073"
id="text1105-6"><tspan
sodipodi:role="line"
id="tspan1103-2"
style="stroke-width:0.264583;fill:#000000"
x="38.879501"
y="40.908073">$x$</tspan></text>
</g>
<path
style="fill:none;stroke:#000000;stroke-width:1;stroke-linecap:butt;stroke-linejoin:miter;stroke-dasharray:none;stroke-opacity:1;marker-end:url(#Arrow2)"
d="M 44.289635,25.433971 H 145.90576"
id="path1414"
sodipodi:nodetypes="cc" />
<text
xml:space="preserve"
style="font-size:8.46667px;line-height:1.25;font-family:sans-serif;text-align:center;text-anchor:middle;stroke-width:0.264583;fill:#000000"
x="92.729836"
y="21.781803"
id="text2092"><tspan
sodipodi:role="line"
id="tspan2090"
style="stroke-width:0.264583;fill:#000000"
x="92.729836"
y="21.781803">$propose(x) / x$</tspan></text>
<text
xml:space="preserve"
style="font-size:8.46667px;line-height:1.25;font-family:sans-serif;text-align:center;text-anchor:middle;stroke-width:0.264583;fill:#000000"
x="166.29887"
y="69.89299"
id="text2092-9"><tspan
sodipodi:role="line"
id="tspan2090-1"
style="stroke-width:0.264583;fill:#000000"
x="166.29887"
y="69.89299">$propose(y) / x$</tspan></text>
<path
style="fill:none;stroke:#000000;stroke-width:1;stroke-linecap:butt;stroke-linejoin:miter;stroke-dasharray:none;stroke-opacity:1;marker-end:url(#Arrow2)"
d="m 155.82329,35.899857 c -8.35129,12.319651 0.54055,24.640898 11.72797,24.072085 8.65403,-0.440005 18.59818,-11.705963 11.8146,-20.570891"
id="path2150"
sodipodi:nodetypes="csc" />
</g>
</svg>

After

Width:  |  Height:  |  Size: 4.7 KiB

File diff suppressed because one or more lines are too long

After

Width:  |  Height:  |  Size: 53 KiB

File diff suppressed because one or more lines are too long

After

Width:  |  Height:  |  Size: 54 KiB

File diff suppressed because one or more lines are too long

After

Width:  |  Height:  |  Size: 56 KiB

Some files were not shown because too many files have changed in this diff Show more