bug 2123: do not try to (confusingly) "repair" missing buckets, just fail #2232
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
|
r+ cbbbb1a |
latobarita
added a commit
that referenced
this pull request
Aug 16, 2019
bug 2123: do not try to (confusingly) "repair" missing buckets, just fail Reviewed-by: MonsieurNicolas
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Outdated suggestions cannot be applied.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.
As mentioned in #2123, the "repair" code is too clever for its own good: it's a sort of wishful-thinking about content addressed buckets that fails to recognize that it has at best a one-in-64 chance of working, and would only ever work when a bucket went missing after being published (and before being replaced with some other local one).
In practice this never happened: the only cases of bucket corruption or missing buckets we've seen in the field were fsync failures and people misconfiguring their BUCKET_DIR, and the "repair" work that kicks in is lost in the error logs, causing the user to totally misinterpret their stalled server as a broken history archive (because the work "can't find" the buckets it's looking for in the archive, and emits an error about that failure-to-download).
This PR just removes the attempt at repair. It is never going to work anyways, and it's better to just be honest and tell the user that with a clear exception. A corrupt bucket dir is actually corrupt and the best repair approach is to kill the node and fast-catch-up to a state before the failure, then let it replay forwards.