-
Notifications
You must be signed in to change notification settings - Fork 3.4k
HBASE-26938 Compaction failures after StoreFileTracker integration #4338
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Closed
Closed
Changes from all commits
Commits
File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
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
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
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
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
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
Oops, something went wrong.
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.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is not needed any more?
Uh oh!
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We can leave it in the API, but the current implementation only set the
writerfield to null, and so the method bodies become empty after that field is converted into a parameter and they no longer have anything to do, so I removed it.If we are going to keep it, we need to use a
Mapinstead ofSetto track the writers, and somehow need to pass a key as a parameter to abort and remove aStoreFileWriterinstance. What should be the key? TheCompactionRequestImpl? I do not think there is a requirement to abort compaction writers in this way. We abort compactions today by interrupting the thread.If
BrokenStoreFileCleanerChorewill not function correctly without this, then it will need modification.I think
BrokenStoreFileCleanerChoresees the same results fromgetCompactionTargetsafter these changes. When the compaction is finished theStoreFileWriterwill be removed from the set in thefinallyblock ofcompact, sogetCompactionTargetswill not include the files being written by that writer after that point, which is the same thing that happened whenresetCompactionWriterwould cause thewriterfield in the previous impl to become null, and also the files being written by that writer would no longer appear ingetCompactionTargetsresults afterward. But the timing has changed, that is true.There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Logically we need this to keep correctness. IIRC, the problem here is that, we can only cleanup the writer instance after we successfully commit the store files to SFT, i.e, after the replaceStoreFile method. That's why we can not just simply remove the writer instance in commitWriter, otherwise there could be data loss, i.e, the BrokenStoreFileCleanerChore may delete the store files which are written just now but have not been added to the SFT yet...
Let me check again if the new implementation can solve the problem.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for that, you know
BrokenStoreFileCleanerChorebest.There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ah actually there is even a problem here in the current code, let me fix it...
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Fixed, at least now we do not remove the writer from the set until after commit.