08:07:26 <yasufum> #startmeeting tacker
08:07:27 <openstack> Meeting started Tue Oct 13 08:07:26 2020 UTC and is due to finish in 60 minutes.  The chair is yasufum. Information about MeetBot at http://wiki.debian.org/MeetBot.
08:07:28 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
08:07:30 <openstack> The meeting name has been set to 'tacker'
08:09:23 <yasufum> takahashi-tsc: are you joining?
08:13:40 <yasufum> I’ve heared that he has something to share about a problem of functional test now he struggling.
08:15:19 <takahashi-tsc> Hi sorry forl ate
08:15:22 <yasufum> hi
08:15:58 <takahashi-tsc> I usually use webchat but now webchat seems be down...
08:16:00 <yasufum> I’ve waited you, for the problem of FT, might because of ceilometer
08:16:12 <takahashi-tsc> Yes I write current status in etherpad
08:16:35 <yasufum> I think it is just one topic from you today.
08:16:42 <yasufum> So, please go ahead
08:16:44 <takahashi-tsc> https://etherpad.opendev.org/p/tacker-meeting
08:17:05 <takahashi-tsc> L469-
08:18:37 <takahashi-tsc> I post a patch to revert ceilometer installation, but after that, POST_FAILURE often occur. I'm not sure the reason is ceilometer, but it works well if remove ceilometer installation.
08:20:09 <takahashi-tsc> Error is journalctl fail. So first I try to reduce log. If failure is not recoverd, we need to remove ceilometer installation agein...
08:20:39 <takahashi-tsc> I'd like to hear opinions or some suggestion from Tacker team.
08:23:13 <yasufum> Before that, could you tell me why you think the reason is the size of logs?
08:23:54 <yasufum> At fiirst?
08:24:48 <yasufum> The error is occured in this job, right? https://opendev.org/openstack/devstack/src/branch/master/roles/export-devstack-journal
08:25:45 <takahashi-tsc> There's few reason... I just assumpt from "failure of jounalctl".
08:26:09 <yasufum> I would like to know there are any other options than removing ceilometer again.
08:26:11 <takahashi-tsc> Probably yes.
08:27:37 <takahashi-tsc> I checked journalctl log with executing "journalctl -o export " and output includes many ceilometer debug logs.
08:28:30 <takahashi-tsc> I'm not sure, but I couldn't think of any other reasons...
08:30:35 <yasufum> Umm, I have roughly understood.
08:32:32 <yasufum> The fix from gmann was effective for our problem actually?
08:32:52 <yasufum> And do you think there are any other problem for ceilometer?
08:35:53 <takahashi-tsc> gmann's fix seems effective for previsou problem, but not effective current problem. It fixed "not instattin Gnocchi" using mariadb, but no other impact.
08:36:19 <takahashi-tsc> I don't know other ceilometer problems...
08:36:59 <yasufum> thanks
08:38:45 <takahashi-tsc> My current plan is written in Etherpad, try reduce log or something today and tomorrow, and if not resolved, I post patch to remove ceilometer installation
08:39:17 <takahashi-tsc> Do you have any comments?
08:40:16 <yasufum> Do you now the way to get the log to understand it is too large before consider to reduce iit?
08:43:51 <yasufum> because we are planning to add some additional tests for the features proposed in wallaby
08:46:27 <yasufum> we have always to consider the size of the log if it is the cause of the problem, because we will add tests for new features.
08:46:50 <takahashi-tsc> I do not know how to check logs in Zuul... I only check the log using my local environment. "jounalctl -o export --since <2 hours ago> > /tmp/file" and check the file.
08:48:15 <yasufum> I mean, we should decide a something limitatioin for that if it is the reason of our problem.
08:50:36 <takahashi-tsc> I see... but the log in my env seems polling log. if it is correct, only way to limit is not output such polling log.
08:50:51 <yasufum> Anyway, do you have any actual plan to reduce the size of the log effectively?
08:51:55 <takahashi-tsc> Set ceilometer's debug flag is current plan. Before that I checked set all debug off but the best is limit only ceilometer and gnocchi debug log
08:52:11 <takahashi-tsc> Set False
08:54:43 <yoshito-ito> For me, it sounds good to try it first. It's better to keep Ceilometer while the use case with it may be not so critical.
08:56:54 <takahashi-tsc> I update etherpad L487-.
08:57:34 <yasufum> thanks
09:00:04 <yasufum> I would like to say, finally, I agree to the strategy of takahashi-tsc.
09:01:23 <yasufum> If we cannot find an effective way other than deactivate ceilometer, we should think to drop it in our release possibly, but want to avoid to do that personally...
09:02:46 <takahashi-tsc> agree. I'd like to keep testing witg ceilometer...
09:04:20 <yasufum> many thanks for your efforts
09:05:47 <yasufum> I would like to close the meeting today if you have no more topics today
09:07:38 <yasufum> Fine, thanks for joining. Bye.
09:07:46 <takahashi-tsc> thanks, bye
09:07:53 <yasufum> #endmeeting