Closed selcukcihan closed 1 year ago
@medikoo ready for review 🙏
@selcukcihan it appears that integreation tests failed after this update: https://github.com/serverless/console/actions/runs/5264703180
@selcukcihan it appears that integreation tests failed after this update: https://github.com/serverless/console/actions/runs/5264703180
Thanks for letting me know, it's a timeout issue (the tests have 120 second total timeout and that was exceeded), I've looked at the action logs to see if I can spot the problem but couldn't come up with any explanation. I'm now trying to reproduce it on my AWS account to see what is causing the timeout. I'll post updates. I suspect it might be a glitch while setting up infra, like waiting for the api to be published or something like that (it's the http api test).
@selcukcihan thanks for looking into that, indeed it could have been glitch on AWS side
I couldn't reproduce it on my account, however looking closer to the logs from the failed action, I can see that the test started at
2023-06-14T08:17:29.2327821Z Python: integration
and the function within the test that took too long was first invoked about 45 seconds later
2023-06-14T08:18:06.277Z i test Invoke function #1 api_endpoint-http-api-v1
And there are no other invocation logs (ie. there is no test Invoke function #2 api_endpoint-http-api-v1
)
That implies either of two cases, with the first one being a strong candidate:
I think there is not much we can do here, except adding a log line in the 404 retry case to ensure when this happens again we know 100% that it's the first case described above that's causing us trouble. What do you think @medikoo ?
I think there is not much we can do here, except adding a log line in the 404 retry case to ensure when this happens again we know 100% that it's the first case described above that's causing us trouble. What do you think @medikoo ?
Adding debug lines that will allow us to have better certainty on what happened is always a good idea
Description
Testing done
Unit/integration tested