QA (public) » openQA Project (public) » openQA Tests (public) » qe-yam

• Changed match_timeout from 120 to 160 and 5 jobs passed.

Please send PR and run 10 at least.

• Now run 20 times. 7 cases were failed to show password dialog.

• PR: https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/20686

• Now run 20 jobs with 180 seconds of waiting time. and 10 jobs failed.

• Run 20 jobs with 240 seconds of waiting time. and 4 out of 20 jobs failed. now I'm checking the code to see the different behavor with the log.

hjluo wrote in #note-11:

• Run 20 jobs with 240 seconds of waiting time. and 4 out of 20 jobs failed. now I'm checking the code to see the different behavor with the log.

I think you can try to reproduce the issue manually, then decide how to act for next step.

leli wrote in #note-12:

hjluo wrote in #note-11:

• Run 20 jobs with 240 seconds of waiting time. and 4 out of 20 jobs failed. now I'm checking the code to see the different behavor with the log.

I think you can try to reproduce the issue manually, then decide how to act for next step.

yes, I'd try, looks like when this issue happened, there's workaroud in the previous screen. Thanks

Did you find any tip how to solve it? (asking as usual as sprint review is near).
It is kind of low prio (and sporadic) compare with other items in the backlog, we could put it on backlog if no more clues, but I'm also fine if you want to continue because you found some way to dig in with the problem. wdyt?

• I did manual test and everyting works, the dialog popped up. now checking log and code for more information.
• I'd like to keep working on it when I was waiting VRs from other tickets.

hjluo wrote in #note-16:

• I did manual test and everyting works, the dialog popped up. now checking log and code for more information.

So this more like a timing issue, please think to increase the RAM and cpu, etc for env effect, more timeout value is also possible.

• I'd like to keep working on it when I was waiting VRs from other tickets.

• Now run 45.2 with QEMURAM=8096 20 jobs. and 4 out of 20 jobs failed for yast2_security.
• Now I'm checking the video to see if there anything abnormal.

• For all the failed jobs, there's one line shows

[2024-12-09T06:50:32.746411Z] [debug] [pid:79637] Check connection to the app by url: http://localhost:39124/v1/widgets

[2024-12-09T06:46:08.217896Z] [info] [pid:14245] yast2_software_management test module started
[2024-12-09T06:46:44.978676Z] [info] [pid:14245] yast2_security test module started
[2024-12-09T06:51:34.121588Z] [debug] [pid:14245] Check connection to the app by url: http://localhost:39119/v1/widgets
[2024-12-09T06:55:58.355376Z] [info] [pid:14245] yast2_lang test module started
[2024-12-09T06:58:41.999747Z] [info] [pid:14245] yast2_users test module started
[2024-12-09T06:59:26.750345Z] [info] [pid:14245] yast2_datetime test module started

for the passed one the output is like:

[2024-12-09T06:46:09.972898Z] [info] [pid:77080] yast2_software_management test module started
[2024-12-09T06:46:47.083504Z] [info] [pid:77080] yast2_security test module started
[2024-12-09T06:52:47.352606Z] [info] [pid:77080] yast2_lang test module started
[2024-12-09T06:55:27.893805Z] [info] [pid:77080] yast2_users test module started
[2024-12-09T06:56:13.262953Z] [info] [pid:77080] yast2_datetime test module started
[2024-12-09T06:56:53.313133Z] [info] [pid:77080] yast2_hostnames test module started
[2024-12-09T06:58:02.315312Z] [info] [pid:77080] yast2_network_settings test module started

• Checked the latest result for RAM=8096 and all failed jobs failed at 3rd time to launch the yast2 security, and the waitmatch wait there for 180 seconds and the password screen still not appeared.

• tried it on x86_64 all 10 jobs passed, looks like this issue just happens on aarch64.

You can try to move this module before yast2_system_settings to have a look.

leli wrote in #note-22:

You can try to move this module before yast2_system_settings to have a look.

OK. tried it on ARM and move security module to the first didn't fix the issue 7 out of 20 jobs failed.

Another thought for this issue, maybe we need do some clean between each test inside the module, suggest to logout and login between each security test.

leli wrote in #note-24:

Another thought for this issue, maybe we need do some clean between each test inside the module, suggest to logout and login between each security test.

• Tried one and looks works and will run 20 jobs to verify it it's works.
• just 4 cases failed with script_out error, looks better now, and changed the code and rerun again.

now all 20 jobs passed, send out PR#20686 for review.

• Lemon reviewed the PR and now changed the code and rerun all VRs.

• Now still in the PR review for newer comments.