Improve the robustness of old ZooKeeper log removal #1040
Merged
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
Motivation:
`OldLogRemover` in `ZooKeeperCommandExecutor` currently catches a
`Throwable` when deleting an old log or its log blocks. However, it has
two issues doing so:
- It doesn't handle an exception that's raised when reading the metadata
of the old log.
- `Throwable` is way too wide exception to catch. Catching a
`KeeperException` whose code is `NONODE` will be enough.
- Note that the failure will only transfer the leadership to other
replica, rather than stopping the whole replication process.
Modifications:
- `OldLogRemover` now catches `KeeperException` whose code is `NONODE`
only.
- An attempt to read a missing log node's metadata is now handled
properly.
- Added more detail to the log messages about missing nodes
- Split `deleteLog()` into `deleteLog()` and `deleteLogBlock()`
Result:
- The leadership is not transferred anymore when `OldLogRemover`
attempts to retrieve a missing log node's metadata, which is not
really a critical issue.
- Instead, the leadership will be transferred when an exception
occurs not because of a missing node.
jrhee17
approved these changes
Oct 8, 2024
Question: Is there any chance that the replica, which receives the leadership, raises the same exception? |
minwoox
pushed a commit
that referenced
this pull request
Dec 5, 2024
…rnings (#1070) **Motivation** With the improvement introduced by #1040, we have been observing logs where CentralDogma attempts to delete logs that don't exist. In order to 1) verify which nodes are being prematurely deleted and 2) to observe trends as to why this may occur, I propose that `LogMeta#toString` be implemented.
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.
Motivation:
OldLogRemoverinZooKeeperCommandExecutorcurrently catches aThrowablewhen deleting an old log or its log blocks. However, it has two issues doing so:Throwableis way too wide exception to catch. Catching aKeeperExceptionwhose code isNONODEwill be enough.Modifications:
OldLogRemovernow catchesKeeperExceptionwhose code isNONODEonly.deleteLog()intodeleteLog()anddeleteLogBlock()Result:
OldLogRemoverattempts to retrieve a missing log node's metadata, which is not really a critical issue.