Closed sysarch-repo closed 3 weeks ago
Thanks for the report.
Obviously something unexpected is happening with the 'ready' attribute.
I did a code review but no obvious cause of such behavior. It is probably somewhere here https://github.com/cnti-testcatalog/testsuite/blob/v1.2.0/src/tasks/workload/microservice.cr#L482-L513
Can you attach complete debug logs from the sig_term_handled execution? If nothing is found there then perhaps I can add some more debug logs and ask you for re-execution.
@martin-mat I believe this issue is closely related to https://github.com/cnti-testcatalog/testsuite/issues/2062 where you provided your initial thoughts. It is the same AUT and there seems to be something wrong with the selection of the objects to be used in the test. I will close this ticket and we will follow up in 2062 with the specialized_ini_system test that is faster to rerun.
Describe the bug The sig_term_handled check is failing due to skipping a pod with container in a not-ready status where both the container and the pod are up and running.
To Reproduce $ cnf-testsuite version CNF TestSuite version: v1.2.0
The issue seems to be specific to a particular pod. When I remove the pod from the AUT, the test PASSes. On the other hand, there is nothing special about this pod among the other pods of the AUT. The only difference in the debug log is that the problematic pod shows restartCount = 1 while others show 0.
My question is why the container statuses JSON is showing ready set to false, while the container status in the YAML pod spec is true. Does it have to do with the restartCount or how exactly does the check work?
Expected behavior If this is a bug, the container status should be correctly picked up by the CNTI testsuite.
Device (please complete the following information): $ uname -a Linux ip-10-0-17-74 6.5.0-1020-aws https://github.com/cnti-testcatalog/testsuite/issues/20~22.04.1-Ubuntu SMP Wed May 1 16:10:50 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux