[dbnode] Fix index flush recovery when previous flush attempts have failed #1574
Conversation
This file contains 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
Codecov Report
@@ Coverage Diff @@
## master #1574 +/- ##
======================================
Coverage 71.8% 71.8%
======================================
Files 946 946
Lines 78002 78002
======================================
Hits 56015 56015
Misses 18343 18343
Partials 3644 3644
Continue to review full report at Codecov.
|
justinjc
reviewed
Apr 19, 2019
| // fileset files has a checkpoint file. | ||
| func (f FileSetFile) HasCheckpointFile() bool { | ||
| func (f FileSetFile) HasCompleteCheckpointFile() bool { | ||
| for _, fileName := range f.AbsoluteFilepaths { | ||
| if strings.Contains(fileName, checkpointFileSuffix) { |
There was a problem hiding this comment.
Probably don't need this line anymore since CompleteCheckpointFileExists does this already.
There was a problem hiding this comment.
True but that will cause an "Invariant error" to be allocated which I'd like to avoid in the case that we're scanning a lot of files.
… of github.com:m3db/m3 into r/fix-index-flush-not-overwriting-half-persisted-files
robskillington
changed the title
added 2 commits
April 19, 2019 17:32
… of github.com:m3db/m3 into r/fix-index-flush-not-overwriting-half-persisted-files
justinjc
approved these changes
Apr 19, 2019
src/dbnode/persist/fs/files.go
Outdated
| type LazyEvalBool uint8 | ||
|
|
||
| const ( | ||
| // EvalNone indicated the boolean has not been evaluated. |
src/dbnode/persist/fs/files.go
Outdated
| @@ -73,14 +73,27 @@ var ( | |||
|
|
|||
| type fileOpener func(filePath string) (*os.File, error) | |||
|
|
|||
| // LazyEvalBool is a boolean is lazily evaluated. | |||
robskillington
deleted the
r/fix-index-flush-not-overwriting-half-persisted-files
branch
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Prior to this change if an index flush failed to complete and write out a checkpoint file, it would leave dirty files that would never be able to be overwritten.
This changes the behavior to only check that the checkpoint file does not exist and is valid, rather than check if a segment file exists or not before writing it.
It also updates all call sites to correctly always validate a checkpoint file rather than just check for existence in other parts of the
persist/fspackage.