Deuxfleurs/garage
48
56
Fork 73
Code Issues 100 Pull requests 13 Releases 30 Wiki Activity

Add compression using zstd #173

Merged
lx merged 8 commits from trinity-1686a/garage:compression into main 2021-12-15 10:26:43 +00:00
Conversation 0 Commits 8 Files changed 8 +18 -5
2 changed files with 18 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files
Showing only changes of commit 75c2766018 - Show all commits

22
doc/book/src/reference_manual/configuration.md
View file

@ -102,12 +102,24 @@ and changing it everywhere at the time), but is not officially supported.
### `compression_level` ### `compression_level`
quentin commented 2021-12-15 09:41:57 +00:00
Review
Copy link

To help people, I propose adding a bit more information here and maybe a link or two?

I found these resources:

The first link says:

The library supports regular compression levels from 1 up to ZSTD_maxCLevel(), which is currently 22. Levels >= 20, labeled --ultra, should be used with caution, as they require more memory. The library also offers negative compression levels, which extend the range of speed vs. ratio preferences. The lower the level, the faster the speed (at the cost of compression).

The CLI explains the 3 different types of compression levels: standard, ultra and fast.

We could add something like that:

Values between 1 (faster compression) and 19 (smaller file) are standard compression levels for zstd. From 20 to 22, compression levels are referred as "ultra" and must be used with extra care as it will use lot of memory. A value of 0 will let zstd choose a default value (currently 3). Finally, zstd has also compression designed to be faster than default compression levels, they range from -1 (smaller file) to -99 (faster compression).

If you do not specify a compression_level entry, garage will set it to 1 for you. With this parameters, zstd consumes low amount of cpu and should work faster than line speed in most situations, while saving some space and intra-cluster
bandwidth.

If you want to totally deactivate zstd in garage, you can pass the special value none. No zstd related code will be called, your chunks will be stored on disk without any processing.

To help people, I propose adding a bit more information here and maybe a link or two? I found these resources: - https://facebook.github.io/zstd/zstd_manual.html - https://github.com/facebook/zstd/blob/d7e17363751974dc1ad10785deb4170b23bee0ec/programs/zstd.1.md - https://github.com/facebook/zstd/blob/9b97fdf74fa9f8adaa557a710b726e2e6966adee/lib/dictBuilder/zdict.c#L768 - https://github.com/facebook/zstd/blob/550410d05d7c7815b1ff417c4cac51153a78785e/lib/zstd.h#L97 - https://github.com/facebook/zstd/blob/38dfc4699e1108d839b3222b6093caaad5befd1c/programs/zstdcli.c#L774-L775 The first link says: > The library supports regular compression levels from 1 up to ZSTD_maxCLevel(), which is currently 22. Levels >= 20, labeled `--ultra`, should be used with caution, as they require more memory. The library also offers negative compression levels, which extend the range of speed vs. ratio preferences. The lower the level, the faster the speed (at the cost of compression). The CLI explains the 3 different types of compression levels: standard, ultra and fast. --- We could add something like that: > Values between `1` (faster compression) and `19` (smaller file) are standard compression levels for zstd. From `20` to `22`, compression levels are referred as "ultra" and must be used with extra care as it will use lot of memory. A value of `0` will let zstd choose a default value (currently `3`). Finally, zstd has also compression designed to be faster than default compression levels, they range from `-1` (smaller file) to `-99` (faster compression). > > If you do not specify a `compression_level` entry, garage will set it to `1` for you. With this parameters, zstd consumes low amount of cpu and should work faster than line speed in most situations, while saving some space and intra-cluster bandwidth. > > If you want to totally deactivate zstd in garage, you can pass the special value `none`. No zstd related code will be called, your chunks will be stored on disk without any processing.
Zstd compression level to use for storing blocks. Can be a positive or negative integer, or Zstd compression level to use for storing blocks.
"none" to disable compression. Default value is 1, which consume low amount of cpu and should
work faster than line speed in most situations, while saving some space and intra-cluster Values between `1` (faster compression) and `19` (smaller file) are standard compression
levels for zstd. From `20` to `22`, compression levels are referred as "ultra" and must be
used with extra care as it will use lot of memory. A value of `0` will let zstd choose a
default value (currently `3`). Finally, zstd has also compression designed to be faster
than default compression levels, they range from `-1` (smaller file) to `-99` (faster
compression).
If you do not specify a `compression_level` entry, garage will set it to `1` for you. With
this parameters, zstd consumes low amount of cpu and should work faster than line speed in
most situations, while saving some space and intra-cluster
bandwidth. bandwidth.
Higher values allow for more compression, at the cost of more cpu usage. Compression is done
synchronously, setting a value too high will add latency to write queries. If you want to totally deactivate zstd in garage, you can pass the special value `'none'`. No
zstd related code will be called, your chunks will be stored on disk without any processing.
Compression is done synchronously, setting a value too high will add latency to write queries.
This value can be different between nodes, compression is done by the node which receive the This value can be different between nodes, compression is done by the node which receive the
API call. API call.

1
script/test-smoke.sh
View file

@ -30,6 +30,7 @@ dd if=/dev/urandom of=/tmp/garage.1.rnd bs=1k count=2 # No multipart, inline sto
dd if=/dev/urandom of=/tmp/garage.2.rnd bs=1M count=5 # No multipart but file will be chunked dd if=/dev/urandom of=/tmp/garage.2.rnd bs=1M count=5 # No multipart but file will be chunked
dd if=/dev/urandom of=/tmp/garage.3.rnd bs=1M count=10 # by default, AWS starts using multipart at 8MB dd if=/dev/urandom of=/tmp/garage.3.rnd bs=1M count=10 # by default, AWS starts using multipart at 8MB
# data of lower entropy, to test compression
dd if=/dev/urandom bs=1k count=2 | base64 -w0 > /tmp/garage.1.b64 dd if=/dev/urandom bs=1k count=2 | base64 -w0 > /tmp/garage.1.b64
dd if=/dev/urandom bs=1M count=5 | base64 -w0 > /tmp/garage.2.b64 dd if=/dev/urandom bs=1M count=5 | base64 -w0 > /tmp/garage.2.b64
dd if=/dev/urandom bs=1M count=10 | base64 -w0 > /tmp/garage.3.b64 dd if=/dev/urandom bs=1M count=10 | base64 -w0 > /tmp/garage.3.b64