garage/src/garage/repair.rs

152 lines
3.6 KiB
Rust
Raw Normal View History

2020-04-23 18:36:12 +00:00
use std::sync::Arc;
use tokio::sync::watch;
2020-07-07 11:59:22 +00:00
use garage_model::garage::Garage;
First implementation of K2V (#293) **Specification:** View spec at [this URL](https://git.deuxfleurs.fr/Deuxfleurs/garage/src/branch/k2v/doc/drafts/k2v-spec.md) - [x] Specify the structure of K2V triples - [x] Specify the DVVS format used for causality detection - [x] Specify the K2V index (just a counter of number of values per partition key) - [x] Specify single-item endpoints: ReadItem, InsertItem, DeleteItem - [x] Specify index endpoint: ReadIndex - [x] Specify multi-item endpoints: InsertBatch, ReadBatch, DeleteBatch - [x] Move to JSON objects instead of tuples - [x] Specify endpoints for polling for updates on single values (PollItem) **Implementation:** - [x] Table for K2V items, causal contexts - [x] Indexing mechanism and table for K2V index - [x] Make API handlers a bit more generic - [x] K2V API endpoint - [x] K2V API router - [x] ReadItem - [x] InsertItem - [x] DeleteItem - [x] PollItem - [x] ReadIndex - [x] InsertBatch - [x] ReadBatch - [x] DeleteBatch **Testing:** - [x] Just a simple Python script that does some requests to check visually that things are going right (does not contain parsing of results or assertions on returned values) - [x] Actual tests: - [x] Adapt testing framework - [x] Simple test with InsertItem + ReadItem - [x] Test with several Insert/Read/DeleteItem + ReadIndex - [x] Test all combinations of return formats for ReadItem - [x] Test with ReadBatch, InsertBatch, DeleteBatch - [x] Test with PollItem - [x] Test error codes - [ ] Fix most broken stuff - [x] test PollItem broken randomly - [x] when invalid causality tokens are given, errors should be 4xx not 5xx **Improvements:** - [x] Descending range queries - [x] Specify - [x] Implement - [x] Add test - [x] Batch updates to index counter - [x] Put K2V behind `k2v` feature flag Co-authored-by: Alex Auvolat <alex@adnab.me> Reviewed-on: https://git.deuxfleurs.fr/Deuxfleurs/garage/pulls/293 Co-authored-by: Alex <alex@adnab.me> Co-committed-by: Alex <alex@adnab.me>
2022-05-10 11:16:57 +00:00
use garage_model::s3::block_ref_table::*;
use garage_model::s3::object_table::*;
use garage_model::s3::version_table::*;
2020-04-24 10:10:01 +00:00
use garage_table::*;
use garage_util::error::Error;
2020-04-23 18:36:12 +00:00
use crate::*;
pub struct Repair {
pub garage: Arc<Garage>,
}
impl Repair {
pub async fn repair_worker(&self, opt: RepairOpt, must_exit: watch::Receiver<bool>) {
if let Err(e) = self.repair_worker_aux(opt, must_exit).await {
warn!("Repair worker failed with error: {}", e);
}
}
async fn repair_worker_aux(
&self,
opt: RepairOpt,
must_exit: watch::Receiver<bool>,
2020-04-23 18:36:12 +00:00
) -> Result<(), Error> {
2021-10-27 08:36:04 +00:00
match opt.what {
RepairWhat::Tables => {
info!("Launching a full sync of tables");
self.garage.bucket_table.syncer.add_full_sync();
self.garage.object_table.syncer.add_full_sync();
self.garage.version_table.syncer.add_full_sync();
self.garage.block_ref_table.syncer.add_full_sync();
self.garage.key_table.syncer.add_full_sync();
}
RepairWhat::Versions => {
info!("Repairing the versions table");
self.repair_versions(&must_exit).await?;
}
RepairWhat::BlockRefs => {
info!("Repairing the block refs table");
self.repair_block_ref(&must_exit).await?;
}
RepairWhat::Blocks => {
info!("Repairing the stored blocks");
self.garage
.block_manager
.repair_data_store(&must_exit)
.await?;
}
RepairWhat::Scrub { tranquility } => {
2021-10-27 08:36:04 +00:00
info!("Verifying integrity of stored blocks");
self.garage
.block_manager
.scrub_data_store(&must_exit, tranquility)
2021-10-27 08:36:04 +00:00
.await?;
}
}
2020-04-23 18:36:12 +00:00
Ok(())
}
async fn repair_versions(&self, must_exit: &watch::Receiver<bool>) -> Result<(), Error> {
let mut pos = vec![];
2022-06-03 10:35:12 +00:00
while *must_exit.borrow() {
let item_bytes = {
let (k, v) = match self.garage.version_table.data.store.get_gt(pos)? {
Some(pair) => pair,
None => break,
};
pos = k.into_vec();
v.into_vec()
};
2020-04-23 18:36:12 +00:00
2022-06-02 21:14:10 +00:00
let version = rmp_serde::decode::from_read_ref::<_, Version>(&item_bytes)?;
if version.deleted.get() {
2020-04-23 18:36:12 +00:00
continue;
}
let object = self
.garage
.object_table
2021-12-14 12:55:11 +00:00
.get(&version.bucket_id, &version.key)
2020-04-23 18:36:12 +00:00
.await?;
let version_exists = match object {
2020-04-26 18:59:17 +00:00
Some(o) => o
.versions()
.iter()
.any(|x| x.uuid == version.uuid && x.state != ObjectVersionState::Aborted),
2021-03-15 14:26:29 +00:00
None => false,
2020-04-23 18:36:12 +00:00
};
if !version_exists {
info!("Repair versions: marking version as deleted: {:?}", version);
self.garage
.version_table
.insert(&Version::new(
version.uuid,
2021-12-14 12:55:11 +00:00
version.bucket_id,
2020-04-23 18:36:12 +00:00
version.key,
true,
))
.await?;
}
}
Ok(())
}
2022-06-02 21:14:10 +00:00
async fn repair_block_ref(&self, must_exit: &watch::Receiver<bool>) -> Result<(), Error> {
let mut pos = vec![];
2020-04-23 18:36:12 +00:00
2022-06-03 10:35:12 +00:00
while *must_exit.borrow() {
let item_bytes = {
let (k, v) = match self.garage.block_ref_table.data.store.get_gt(pos)? {
Some(pair) => pair,
None => break,
};
pos = k.into_vec();
v.into_vec()
};
2022-06-02 21:14:10 +00:00
let block_ref = rmp_serde::decode::from_read_ref::<_, BlockRef>(&item_bytes)?;
if block_ref.deleted.get() {
2020-04-23 18:36:12 +00:00
continue;
}
let version = self
.garage
.version_table
.get(&block_ref.version, &EmptyKey)
.await?;
2021-03-15 14:26:29 +00:00
// The version might not exist if it has been GC'ed
let ref_exists = version.map(|v| !v.deleted.get()).unwrap_or(false);
2020-04-23 18:36:12 +00:00
if !ref_exists {
info!(
"Repair block ref: marking block_ref as deleted: {:?}",
block_ref
);
self.garage
.block_ref_table
.insert(&BlockRef {
block: block_ref.block,
version: block_ref.version,
deleted: true.into(),
2020-04-23 18:36:12 +00:00
})
.await?;
}
}
Ok(())
}
}