-
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
roachtest: scaledata/filesystem_simulator/nodes=6 failed #51215
Comments
Duplicate of #51208. |
Actually I'll keep this open since it's on master. |
(roachtest).scaledata/filesystem_simulator/nodes=6 failed on master@542c83e7db1694208e09c7fea7d350db4993b897:
More
Artifacts: /scaledata/filesystem_simulator/nodes=6 See this test on roachdash |
(roachtest).scaledata/filesystem_simulator/nodes=6 failed on master@12b58af06d9104bc585bc0773bbe29d68ff2d826:
More
Artifacts: /scaledata/filesystem_simulator/nodes=6
See this test on roachdash |
(roachtest).scaledata/filesystem_simulator/nodes=6 failed on master@161d68e8c2149d401738d2c52cf70071a1c3b610:
More
Artifacts: /scaledata/filesystem_simulator/nodes=6
See this test on roachdash |
Things fail after an unexpected query execution cancellation error.
There's nothing in the client test code (https://github.com/cockroachdb/rksql) as far as I can tell that would send forward query cancellations. I don't think these are statement timeouts either, the statements are tiny, terminate quickly, and I think they would propagate a different error tag (haven't verified).
The discussion in #51375 (comment) and #51518 (comment) seem relevant. |
I'll note that this failure started after we resolved #50687, but if I had to guess I'd guess that #50687 was just masking this particular error mode (and may have been in part due to |
I looked into this. The actual log output is this:
This tells us 3 things:
I don't see explicit CANCEL requests, so I go to fish in the node logs to see more details. Unfortunately I don't see these query timeouts being canceled in logs. I would suggest a couple of "next actions" here:
@asubiotto @yuzefovich do you want to create an issue just for items 1 and 2? |
sorry my last sentence was meant to be "For items 1 and 2" not "2 and 3" |
In all my prior attempts at reproing it I was running local builds, just tried a couple (by that I mean literally 2) of non local runs and was able to run into it each time.
This is on your run-of-the-mill n4cpu4s. I'll optimistically mark this as "reproducible" and hand it over to @yuzefovich and @asubiotto. Seems it's a pretty pervasive problem (#51647). |
(roachtest).scaledata/filesystem-simulator/nodes=6 failed on master@e9a4f83e3eee59510f97db2c6e0df9b57cf6b944:
More
Artifacts: /scaledata/filesystem-simulator/nodes=6 See this test on roachdash |
I think my original statement that we could see "query canceled" errors because cockroach/pkg/sql/conn_executor_exec.go Lines 192 to 197 in 89dda79
we return a different query execution canceled due to statement timeout error. And I'm not aware of any other system timeouts in SQL layer, so I don't think there is anything to do for this item.
This is similar to point 1 - I'm only aware of @knz do you agree with my reasoning? For more context, we have been seeing "query canceled" errors because of a bug in the vectorized engine which I have a PR (#51772) to fix. I'm hoping that once that goes in, these "query canceled" errors will go away. |
(roachtest).scaledata/filesystem-simulator/nodes=6 failed on master@b8a50cc4d062293915969cdc83e3ec4d057cede5:
More
Artifacts: /scaledata/filesystem-simulator/nodes=6 See this test on roachdash |
(roachtest).scaledata/filesystem-simulator/nodes=6 failed on master@bfa6307c292ef4dfed4a53cb99f506e6dab26533:
More
Artifacts: /scaledata/filesystem-simulator/nodes=6 See this test on roachdash |
(roachtest).scaledata/filesystem-simulator/nodes=6 failed on master@a16eb55ed96239dcd288aa1c2f80f306559f0f0b:
More
Artifacts: /scaledata/filesystem-simulator/nodes=6 See this test on roachdash |
(roachtest).scaledata/filesystem-simulator/nodes=6 failed on master@3edbe4aeb3c7300e6690cb2222a8d5c01e920bf4:
More
Artifacts: /scaledata/filesystem-simulator/nodes=6 See this test on roachdash |
#51772 landed yesterday afternoon, and there wasn't any failure from the nightly run yesterday, so closing 🤞 Thanks @yuzefovich! |
(roachtest).scaledata/filesystem_simulator/nodes=6 failed on master@1b5d070c93375d3e14c146241e8bafde349529bd:
More
Artifacts: /scaledata/filesystem_simulator/nodes=6
Related:
See this test on roachdash
powered by pkg/cmd/internal/issues
The text was updated successfully, but these errors were encountered: