Cleanup Concurrent RepositoryData Loading#48329
Merged
original-brownbear merged 5 commits intoelastic:masterfrom Oct 23, 2019
original-brownbear:48122
Merged
Cleanup Concurrent RepositoryData Loading#48329original-brownbear merged 5 commits intoelastic:masterfrom original-brownbear:48122
original-brownbear merged 5 commits intoelastic:masterfrom
original-brownbear:48122
Conversation
The loading of `RepositoryData` is not an atomic operation. It uses a list + get combination of calls. This lead to accidentally returning an empty repository data for generations >=0 which can never not exist unless the repository is corrupted. In the test #48122 (and other SLM tests) there was a low chance of running into this concurrent modification scenario and the repository actually moving two index generations between listing out the index-N and loading the latest version of it. Since we only keep two index-N around at a time this lead to unexpectedly absent snapshots in status APIs. Fixing the behavior to be more resilient is non-trivial but in the works. For now I think we should simply throw in this scenario. This will also help prevent corruption in the unlikely event but possible of running into this issue in a snapshot create or delete operation on master failover on a repository like S3 which doesn't have the "no overwrites" protection on writing a new index-N. Fixes #48122
Collaborator
|
Pinging @elastic/es-distributed (:Distributed/Snapshot/Restore) |
Contributor
Author
|
I didn't yet get around to a full test run here, so if this doesn't yet go green please hold off reviewing till I can fix the tests I may have missed :) |
| public RepositoryData getRepositoryData() { | ||
| try { | ||
| return getRepositoryData(latestIndexBlobId()); | ||
| } catch (NoSuchFileException ex) { |
Contributor
Author
There was a problem hiding this comment.
There is no good reason to catch here or below. We break out on generation -1 and return empty data so whenever we fail to find data for gen. >= 0 it's a problem and we mustn't ignore it.
Contributor
Author
|
Jup, this is not entirely unsurprisingly failing some REST tests: -> WIP EDIT: #46250 fixed the tests here sort of by accident by changing the order of shard and root level meta updates :) -> should be good to review now |
tlrx
approved these changes
Oct 23, 2019
| repositoryData = RepositoryData.snapshotsFromXContent(parser, indexGen); | ||
| } | ||
| return repositoryData; | ||
| } catch (NoSuchFileException ex) { |
Member
There was a problem hiding this comment.
nit: while we are at it, there are unneeded Long.toString() and local variable repositoryData usages
Contributor
Author
There was a problem hiding this comment.
I can't find it right now, but others disagreed with removing the Long.toString in another PR :( But yea let's clean up the redundant locals :)
Contributor
Author
|
Thanks Tanguy! |
This was referenced Oct 23, 2019
original-brownbear
added a commit
that referenced
this pull request
Oct 24, 2019
This was referenced Nov 2, 2019
original-brownbear
added a commit
that referenced
this pull request
Nov 2, 2019
The loading of `RepositoryData` is not an atomic operation. It uses a list + get combination of calls. This lead to accidentally returning an empty repository data for generations >=0 which can never not exist unless the repository is corrupted. In the test #48122 (and other SLM tests) there was a low chance of running into this concurrent modification scenario and the repository actually moving two index generations between listing out the index-N and loading the latest version of it. Since we only keep two index-N around at a time this lead to unexpectedly absent snapshots in status APIs. Fixing the behavior to be more resilient is non-trivial but in the works. For now I think we should simply throw in this scenario. This will also help prevent corruption in the unlikely event but possible of running into this issue in a snapshot create or delete operation on master failover on a repository like S3 which doesn't have the "no overwrites" protection on writing a new index-N. Fixes #48122
original-brownbear
added a commit
that referenced
this pull request
Nov 2, 2019
The loading of `RepositoryData` is not an atomic operation. It uses a list + get combination of calls. This lead to accidentally returning an empty repository data for generations >=0 which can never not exist unless the repository is corrupted. In the test #48122 (and other SLM tests) there was a low chance of running into this concurrent modification scenario and the repository actually moving two index generations between listing out the index-N and loading the latest version of it. Since we only keep two index-N around at a time this lead to unexpectedly absent snapshots in status APIs. Fixing the behavior to be more resilient is non-trivial but in the works. For now I think we should simply throw in this scenario. This will also help prevent corruption in the unlikely event but possible of running into this issue in a snapshot create or delete operation on master failover on a repository like S3 which doesn't have the "no overwrites" protection on writing a new index-N. Fixes #48122
This was referenced Feb 3, 2020
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The loading of
RepositoryDatais not an atomic operation.It uses a list + get combination of calls.
This lead to accidentally returning an empty repository data
for generations >=0 which can never not exist unless the repository
is corrupted.
In the test #48122 (and other SLM tests) there was a low chance of
running into this concurrent modification scenario and the repository
actually moving two index generations between listing out the
index-N and loading the latest version of it. Since we only keep
two index-N around at a time this lead to unexpectedly absent
snapshots in status APIs.
Fixing the behavior to be more resilient is non-trivial but in the works.
For now I think we should simply throw in this scenario. This will also
help prevent corruption in the unlikely event but possible of running into this
issue in a snapshot create or delete operation on master failover on a
repository like S3 which doesn't have the "no overwrites" protection on
writing a new index-N.
I would suggest back porting this all the way since it can theoretically have
repository corrupting effects on S3.
Fixes #48122