Simplify and optimize deduplication of RepositoryData for a non-caching repository instance #91851
Conversation
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
…ng repository instance This makes use of the new deduplicator infrastructure to move to more efficient deduplication mechanics. The existing solution hardly ever deduplicated because it would only deduplicate after the repository entered a consistent state. The adjusted solution is much simpler, in that it simply deduplicates such that only a single loading of `RepositoryData` will ever happen at a time, fixing memory issues from massively concurrent loading of the repo data as described in elastic#89952. closes elastic#89952
original-brownbear
added
>bug
:Distributed Coordination/Snapshot/Restore
elasticsearchmachine
added
the
Team:Distributed (Obsolete)
Collaborator
|
Pinging @elastic/es-distributed (Team:Distributed) |
Collaborator
|
Hi @original-brownbear, I've created a changelog YAML for you. |
DaveCTurner
reviewed
Nov 23, 2022
Contributor
DaveCTurner
left a comment
DaveCTurner
left a comment
There was a problem hiding this comment.
Nice simplification. I left one small comment.
| this.basePath = basePath; | ||
| this.maxSnapshotCount = MAX_SNAPSHOTS_SETTING.get(metadata.settings()); | ||
| this.repoDataDeduplicator = new ResultDeduplicator<>(threadPool.getThreadContext()); | ||
| this.repoDataLoadDeduplicator = new SingleResultDeduplicator<>(threadPool.getThreadContext(), this::doGetRepositoryData); |
Contributor
There was a problem hiding this comment.
I think SingleResultDeduplicator risks a stack overflow exception if the action doesn't fork, which is the case here. Can we fork to SNAPSHOT_META here rather than when calling repoDataLoadDeduplicator::execute?
Contributor
Author
There was a problem hiding this comment.
++ I pushed below commit, thanks for catching this! Also simplifies the call-sites nicely :)
DaveCTurner
approved these changes
Nov 23, 2022
| repoData.getGenId() | ||
| ) | ||
| ) | ||
| repoDataLoadDeduplicator.execute( |
Contributor
Author
|
Thanks David! |
original-brownbear
added a commit
to original-brownbear/elasticsearch
that referenced
this pull request
Nov 23, 2022
…ng repository instance (elastic#91851) This makes use of the new deduplicator infrastructure to move to more efficient deduplication mechanics. The existing solution hardly ever deduplicated because it would only deduplicate after the repository entered a consistent state. The adjusted solution is much simpler, in that it simply deduplicates such that only a single loading of `RepositoryData` will ever happen at a time, fixing memory issues from massively concurrent loading of the repo data as described in elastic#89952. closes elastic#89952
elasticsearchmachine
pushed a commit
that referenced
this pull request
Jan 3, 2023
…-caching repository instance (#91851) (#91866) * Simplify and optimize deduplication of RepositoryData for a non-caching repository instance (#91851) This makes use of the new deduplicator infrastructure to move to more efficient deduplication mechanics. The existing solution hardly ever deduplicated because it would only deduplicate after the repository entered a consistent state. The adjusted solution is much simpler, in that it simply deduplicates such that only a single loading of `RepositoryData` will ever happen at a time, fixing memory issues from massively concurrent loading of the repo data as described in #89952. closes #89952 * fix compile
Contributor
|
Would it be safe to backport this to 7.17 too? We definitely see #89952 affecting them, and I don't see any obvious reasons not to apply this there. |
Contributor
Author
|
Yea it should be fine :) let me try .. |
Contributor
Author
|
#92661 worked |
original-brownbear
added a commit
to original-brownbear/elasticsearch
that referenced
this pull request
Jan 4, 2023
…-caching repository instance (elastic#91851) (elastic#91866) * Simplify and optimize deduplication of RepositoryData for a non-caching repository instance (elastic#91851) This makes use of the new deduplicator infrastructure to move to more efficient deduplication mechanics. The existing solution hardly ever deduplicated because it would only deduplicate after the repository entered a consistent state. The adjusted solution is much simpler, in that it simply deduplicates such that only a single loading of `RepositoryData` will ever happen at a time, fixing memory issues from massively concurrent loading of the repo data as described in elastic#89952. closes elastic#89952 * fix compile
elasticsearchmachine
pushed a commit
that referenced
this pull request
Jan 4, 2023
…-caching repository instance (#91851) (#91866) (#92661) * Simplify and optimize deduplication of RepositoryData for a non-caching repository instance (#91851) This makes use of the new deduplicator infrastructure to move to more efficient deduplication mechanics. The existing solution hardly ever deduplicated because it would only deduplicate after the repository entered a consistent state. The adjusted solution is much simpler, in that it simply deduplicates such that only a single loading of `RepositoryData` will ever happen at a time, fixing memory issues from massively concurrent loading of the repo data as described in #89952. closes #89952 * fix compile
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
This makes use of the new deduplicator infrastructure to move to more efficient deduplication mechanics.
The existing solution hardly ever deduplicated because it would only deduplicate after the repository entered a consistent state. The adjusted solution is much simpler, in that it simply deduplicates such that only a single loading of
RepositoryDatawill ever happen at a time, fixing memory issues from massively concurrent loading of the repo data as described in #89952.closes #89952 (this should be all that's needed to fix the memory issue in practice, judging by heap dumps as the problem really just is the initial set of snapshots going wild with concurrent loading of repo data on the huge snapshot-meta pool)