-
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
kv/kvserver: TestReplicateQueueUpAndDownReplicateNonVoters failed #75135
Labels
branch-master
Failures and bugs on the master branch.
C-test-failure
Broken test (automatically or manually discovered).
O-robot
Originated from a bot.
Comments
cockroach-teamcity
added
branch-master
Failures and bugs on the master branch.
C-test-failure
Broken test (automatically or manually discovered).
O-robot
Originated from a bot.
labels
Jan 19, 2022
cc @irfansharif this is probably zonecfg fallout. |
@arulajmani with #75072, were we able to repro the original failure reliably? That PR looks correct, but perhaps there's something additional we missed. Mind taking a look? |
irfansharif
added a commit
to irfansharif/cockroach
that referenced
this issue
Jan 20, 2022
Fixes cockroachdb#75135. This test asserted on span configs applying to a scratch range. When stressing, it appeared that some time we were not seeing the scratch range adopt the prescribed number of voters/non-voters. Staring at the test itself, we were only nudging the replication queues for the first node in the three node test. It's possible for the scratch range to have been housed on a node other than the first; this commit makes it so that the test nudges queues on all nodes. For good measure, lets also ensure that the split queues process everything, ditto for the snapshot queues. To repro: dev test pkg/kv/kvserver \ -f TestReplicateQueueUpAndDownReplicateNonVoters \ -v --show-logs --timeout 2m --stress Release note: None
craig bot
pushed a commit
that referenced
this issue
Jan 20, 2022
75169: vendor: pull in latest version of `stress` r=rail a=rickystewart Pull in the latest version of `stress` including these changes: ``` 43d99a9 Merge pull request #13 from cockroachdb/bazelsharding 01690a1 stress: add `-bazel` support, support for sharding artifacts ``` Release note: None 75194: cli: stop ignoring user arg in insecure mode r=otan,knz a=rafiss fixes #74704 Release note (bug fix): The --user argument is no longer ignored when using `cockroach sql` in --insecure mode. 75212: kvserver: de-flake TestReplicateQueueUpAndDownReplicateNonVoters r=irfansharif a=irfansharif Fixes #75135. This test asserted on span configs applying to a scratch range. When stressing, it appeared that some time we were not seeing the scratch range adopt the prescribed number of voters/non-voters. Staring at the test itself, we were only nudging the replication queues for the first node in the three node test. It's possible for the scratch range to have been housed on a node other than the first; this commit makes it so that the test nudges queues on all nodes. For good measure, lets also ensure that the split queues process everything, ditto for the snapshot queues. To repro: dev test pkg/kv/kvserver \ -f TestReplicateQueueUpAndDownReplicateNonVoters \ -v --show-logs --timeout 2m --stress Release note: None 75218: backupccl: disable span configs for full cluster restore jobs test r=irfansharif a=RichardJCai Release note: None Related #75216 75220: kvserver: disable sendWithRangeID call stack r=knz a=tbg Sadly, this on longer embeds the RangeID in the stack trace (likely culprit: Go adopting a register-based calling convention) Instead, you get garbage values that often are obviously garbage, but this may not always be true. We want to avoid being misled, so for now remove the rangeID here and explain when it can come back. See: https://cockroachlabs.slack.com/archives/G01G8LK77DK/p1641478596004700 Release note: None Co-authored-by: Ricky Stewart <[email protected]> Co-authored-by: Rafi Shamim <[email protected]> Co-authored-by: irfan sharif <[email protected]> Co-authored-by: richardjcai <[email protected]> Co-authored-by: Tobias Grieger <[email protected]>
gtr
pushed a commit
to gtr/cockroach
that referenced
this issue
Jan 24, 2022
Fixes cockroachdb#75135. This test asserted on span configs applying to a scratch range. When stressing, it appeared that some time we were not seeing the scratch range adopt the prescribed number of voters/non-voters. Staring at the test itself, we were only nudging the replication queues for the first node in the three node test. It's possible for the scratch range to have been housed on a node other than the first; this commit makes it so that the test nudges queues on all nodes. For good measure, lets also ensure that the split queues process everything, ditto for the snapshot queues. To repro: dev test pkg/kv/kvserver \ -f TestReplicateQueueUpAndDownReplicateNonVoters \ -v --show-logs --timeout 2m --stress Release note: None
23 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Labels
branch-master
Failures and bugs on the master branch.
C-test-failure
Broken test (automatically or manually discovered).
O-robot
Originated from a bot.
kv/kvserver.TestReplicateQueueUpAndDownReplicateNonVoters failed with artifacts on master @ 912964e02ddd951c77d4f71981ae18b3894e9084:
Help
See also: How To Investigate a Go Test Failure (internal)
Parameters in this failure:
TAGS=deadlock
GOFLAGS=-parallel=2
This test on roachdash | Improve this report!
The text was updated successfully, but these errors were encountered: