-
Notifications
You must be signed in to change notification settings - Fork 3.9k
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/kvnemesis: TestKVNemesisSingleNode failed #118005
Comments
kv/kvnemesis.TestKVNemesisSingleNode failed on master @ 3c837c2a86165188a26e8629b7e21b7ae0fb56ae: Fatal error:
Stack:
Log preceding fatal error
Parameters:
Same failure on other branches
|
Seems to be stuck waiting in the lock table for over 4 minutes:
|
This comment was marked as off-topic.
This comment was marked as off-topic.
Looks like there's some sort of deadlock here, given we're waiting in the |
I've been running this on my GCE worker for more than an hour and I'm yet to see a failure. |
@lyang24 that PR also only affects Another hour and a half on my GCE worker and nothing. I'm beginning to wonder if there's something specific to EngFlow going on here. |
Previously, kvnemesis used os.TempDir() to write the various debug files (including the repro steps) to a temp dir. When the test failed on EngFlow, the temp dir was not included in output.zip, which made tests hard to investigate. This patch uses datapathutils.DebuggableTempDir() instead. If the test is running locally, the behavior is the same as os.TempDir(). If the test is running remotely, it will write the debug files to TEST_UNDECLARED_OUTPUTS_DIR and Bazel will package them up into outputs.zip. Informs: cockroachdb#118005 Release note: None
118317: kvnemesis: write debug files using DebuggableTempDir r=rickystewart a=miraradeva Previously, kvnemesis used os.TempDir() to write the various debug files (including the repro steps) to a temp dir. When the test failed on EngFlow, the temp dir was not included in output.zip, which made tests hard to investigate. This patch uses datapathutils.DebuggableTempDir() instead. If the test is running locally, the behavior is the same as os.TempDir(). If the test is running remotely, it will write the debug files to TEST_UNDECLARED_OUTPUTS_DIR and Bazel will package them up into outputs.zip. Informs: #118005 Release note: None Co-authored-by: Mira Radeva <[email protected]>
kv/kvnemesis.TestKVNemesisSingleNode failed on master @ 0baf22a03d5f55e2611701bc723e3e0b713ab051: Fatal error:
Stack:
Log preceding fatal error
Parameters:
|
If need be, feel free to get a |
116958: rpc: move system ranges to system RPC class r=lunevalex a=lunevalex Move all the ranges between /Min and /System/tsd to use the default RPC class. This will allow for isolation from network congestion for all system ranges, which crucial for the stability of the system. Fixes: #111239 Release note: None 118555: workflows: run UI lint and test in experimental github actions build r=rail a=rickystewart Epic: [CRDB-8308](https://cockroachlabs.atlassian.net/browse/CRDB-8308) Release note: None 118581: roachtest: stop ignoring activerecord failures r=rafiss a=rafiss The adapter has been stabilized, so we should enable this test again. fixes #108938 Release note: None 118659: tests: use `test.Pool` instead of `Pool` r=rail a=rickystewart This tells Bazel to use the pool only for test actions instead of the compile action associated with each test. Epic: CRDB-8308 Release note: None 118673: kv: mark kvnemesis tests as "large" sized r=nvanbenschoten a=arulajmani We've recently seen these time out exclusively on eng flow. In all those instances, we can see the test is making some progress from the stack traces -- it's slow though. We mark KVNemesis tests as large, which in turn bumps their timeout in CI. Closes #118624 Closes #118005 Release note: None 118675: ui: remove warning when auto refresh enable r=maryliag a=maryliag The warning being displayed about old active executions was not bein properly removed when the auto refresh was turned back on. This commit fixes this for both Statements and Transactions pages, on Active Executions. Fixes CRDB-35837 https://www.loom.com/share/76b57eba17ab44758fe81f178f07fecd Release note (ui change): Properly remove warning of old date on Active Executions when auto refresh is enabled. Co-authored-by: Alex Lunev <[email protected]> Co-authored-by: Ricky Stewart <[email protected]> Co-authored-by: Rafi Shamim <[email protected]> Co-authored-by: Arul Ajmani <[email protected]> Co-authored-by: maryliag <[email protected]>
We've recently seen these time out exclusively on eng flow. In all those instances, we can see the test is making some progress from the stack traces -- it's slow though. We mark KVNemesis tests as large, which in turn bumps their timeout in CI. Closes cockroachdb#118624 Closes cockroachdb#118005 Release note: None
kv/kvnemesis.TestKVNemesisSingleNode failed on master @ f740c076ec0d92972a47cf6633cf065c8c98678f:
Fatal error:
Stack:
Log preceding fatal error
Parameters:
attempt=1
run=30
shard=1
Help
See also: How To Investigate a Go Test Failure (internal)
Same failure on other branches
This test on roachdash | Improve this report!
Jira issue: CRDB-35451
The text was updated successfully, but these errors were encountered: