Phil/pruning paranoia #363
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
If someone accidentally runs `shards prune` with a selector that doesn't include all the shards that use a forked recovery log, it's possible for it to try to prune the current fragment for a journal. This adds a check for that condition so that we can provide visibility and stop the prune operation. Attempting to prune an unpersisted fragment would fail anyway, but this at least gives a clearer error message.
Adds a lot more information to the logs when fragments are deleted. The intent is to run with JSON logs, to enable automated audits and analysis of prune operations. Because we're logging out so much more information as part of the logs, I also turned down the log level on some of the other messages. I also added in some additional warning logs for conditions that should never be encountered.
This allows recovery log pruning to continue after encountering an error removing a fragment. We now operate Gazette clusters that use a variety of different storage buckets, and it seems unavoidable that some of them might have permissions misconfigured or return an error for some other reason. In that case, we'll now log a warning and continue the prune operation. Gazctl will still exit non-zero if it has encountered any errors removing fragments, to ensure it never fails silently.
# for free
to join this conversation on GitHub.
Already have an account?
# 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Outdated suggestions cannot be applied.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.
Rolls up a few minor improvements to
gazctl shards prune. Individual commit messages have more details, but the main goal is to add more detailed audit logging, and to continue pruning after failing to remove a fragment.I've been using this script to compare the audit logs produced by this, to try to detect any journals or fragments that were removed in an earlier prune operation, but not in a subsequent one. So far, I've not found any.
This change is