-
Notifications
You must be signed in to change notification settings - Fork 4.8k
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
Test flake: //test/extensions/filters/http/ext_proc:ext_proc_integration_test #36041
Comments
@cainelli Do you mind taking a look at this? Thank you! I think tracing related feature was added by your change |
oh sorry about that. I will take a look into it this week. |
@tyxia the failure is a timeout and the tracing failure is a side effect. I don't see how such simple request would take more than 5s to run.
|
envoy/test/integration/http_integration.cc Lines 555 to 556 in 7a7df5d
update: nvm is 5s envoy/test/extensions/filters/http/ext_proc/ext_proc_integration_test.cc Lines 280 to 281 in 7a7df5d
|
sorry about the dup but can you please have a look at the test? Flaky tests are problematic for everyone else trying to submit to Envoy! |
@cainelli Thanks for spending effort reducing the flakiness. In the past, we have test (as you linked in slack) that have larger then 5s timeout. However, that is because I am not sure if the flakiness is because tracing will take a bit more time but 5s should be sufficient here. Or maybe because ext_proc_integration test has grown very big now. Do you happen to know what is the flakiness rate? |
if you can't repro the flake (per slack) one thing you can do is add a LogLevelSetter in that test such that CI logs more information when it flakes. then next time we see a failure you'll have more information. |
One thing I've found often helps if you can't repro a flake is to run |
Thank you all for the context and tips.
I did tried that with various combinations to stress during the test but did not have any luck reproducing it.
I will try this path moving forward (#36583). |
Commit Message: Increase log verbosity and add tracing logs to help diagnose potential flakiness in the GetAndCloseStreamWithTracing test case. Additional Description: As per discussed in #36041 (comment) Risk Level: Low --------- Signed-off-by: Fernando Cainelli <fernando.cainelli-external@getyourguide.com>
https://dev.azure.com/cncf/envoy/_build/results?buildId=179508&view=logs&j=4930ecaf-18f4-5b3c-dea3-309729c3b3ae&t=573d8780-d7b9-52e3-b4e0-a89886b0b9ff&l=3840
@tyxia
The text was updated successfully, but these errors were encountered: