-
Notifications
You must be signed in to change notification settings - Fork 3.8k
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
release-23.2: storage: disable crashing for single delete callbacks #116945
release-23.2: storage: disable crashing for single delete callbacks #116945
Conversation
As elaborated in code comments, these can be false positives caused by delete-only compactions. The metrics are not disabled, under the rare chance that we see a problem in a cluster and one of the metrics (if it happens to not be a false positive) gives us a hint about the cause of the problem. For the same reason, we log such callbacks every 5 minutes, so we can correlate the key in such logs with an actual problem. False positives should be rare, especially for the invariant violation callback. Informs #115881 Informs #114421 Epic: none Release note (ops change): The cluster settings storage.single_delete.crash_on_invariant_violation.enabled and storage.single_delete.crash_on_ineffectual.enabled are disabled and must not be enabled.
5c23bdb
to
8dfe2cd
Compare
Thanks for opening a backport. Please check the backport criteria before merging:
If your backport adds new functionality, please ensure that the following additional criteria are satisfied:
Also, please add a brief release justification to the body of your PR to justify this |
It looks like your PR touches production code but doesn't add or edit any test code. Did you consider adding tests to your PR? 🦉 Hoot! I am a Blathers, a bot for CockroachDB. My owner is dev-inf. |
ftr, #116897 was the v23.2.0 backport. |
Backport 1/1 commits from #116889 on behalf of @sumeerbhola.
/cc @cockroachdb/release
As elaborated in code comments, these can be false positives caused by delete-only compactions. The metrics are not disabled, under the rare chance that we see a problem in a cluster and one of the metrics (if it happens to not be a false positive) gives us a hint about the cause of the problem. For the same reason, we log such callbacks every 5 minutes, so we can correlate the key in such logs with an actual problem. False positives should be rare, especially for the invariant violation callback.
Informs #115881
Informs #114421
Epic: none
Release note (ops change): The cluster settings
storage.single_delete.crash_on_invariant_violation.enabled and storage.single_delete.crash_on_ineffectual.enabled are disabled and must not be enabled.
Release justification: Removes flawed assertion that could crash nodes.