Flaky Test: solana-local-cluster::local_cluster test_duplicate_shreds_switch_failure

[anza-team]

AUTO-GENERATED. DO NOT EDIT.

📝 Buildkite Analytics

[steviez]

Ran into this one 5 times in a row; might poke around a little or poke those who know more than me about this test

[AshwinSekar]

There is flakiness in this test surrounding the gossip propagation of the duplicate proof not be deterministic. I will fix this as part of some upcoming work I am doing to improve the duplicate proof.

[steviez]

I'm not sure we should close the issue just yet - @AshwinSekar's linked PR just ignores the test, but I don't think we have actually fixed it yet. If it has been properly fixed, we should link the PR that closes it to leave a paper-trail. In any case, going to reopen so Ashwin can chime in if there are any updates on this

[yihau]

I thought it was ignored, so it's not counted as a flaky test in our current CI environment, but yeah, keeping it open truly helps us for easy tracking. let's keep it open!

[steviez]

I thought it was ignored, so it's not counted as a flaky test in our current CI environment,

Correct, it is currently being ignored. So yes, I guess I would agree that it is technically not flaky in its' current state 🤣

Please correct me if I'm mistaken Ashwin, but I'm inclined to think there is value in making the test more deterministic and re-enabling. If we close the issue, I think it would be likely that we forget about it and never start running it again.

I see the issue was opened by a bot. If you're using issues created by that account as a filter, I think we could probably close and reopen the issue with a new title "Fix and re-enable test_duplicate_shreds_switch_failure", especially since we'll probably have to do that to move it to the new repo soon

[AshwinSekar]

I added it to the list of tests we need to fix here #34857