16:01:27 <gthiemonge> #startmeeting Octavia
16:01:27 <opendevmeet> Meeting started Wed Feb  8 16:01:27 2023 UTC and is due to finish in 60 minutes.  The chair is gthiemonge. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:27 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:27 <opendevmeet> The meeting name has been set to 'octavia'
16:01:35 <johnsom> o/
16:01:40 <tweining> o/
16:01:44 <gthiemonge> Hi Folks
16:03:14 <matfechner> o/
16:03:19 <oschwart> o/
16:03:47 <gthiemonge> #topic Announcements
16:04:07 <johnsom> octavia-lib for Antelope released
16:04:11 <gthiemonge> Next week is Antelope-3 milestone: Feature Freeze
16:04:15 <gthiemonge> +1
16:04:27 <johnsom> Yep, and client releases
16:04:27 <gthiemonge> I'll focus on reviewing feature patches
16:04:54 <gthiemonge> there's only one patch for the client
16:05:10 <gthiemonge> reminder, the review priority list is:
16:05:13 <gthiemonge> #link https://etherpad.opendev.org/p/octavia-priority-reviews
16:06:07 <johnsom> I have +2'd that already. Maybe tweining can have a look
16:07:04 <tweining> ok
16:07:26 <gthiemonge> any other announcement folks?
16:09:48 <gthiemonge> #topic CI Status
16:10:09 <gthiemonge> we fixed issues with grenade and pylint this week, I think all the fixes are merged now
16:10:16 <gthiemonge> (master and stable branches)
16:11:14 <gthiemonge> another topic:
16:11:46 <gthiemonge> froyo is adding new tests (missing tests) to test_traffic_ops in octavia-tempest-plugin
16:12:25 <gthiemonge> it increases the duration of the scenario jobs, the c9s job seems to time out really often now
16:12:57 <gthiemonge> I tried to split the class into 2 classes but c9s still fails from time to time
16:13:23 <gthiemonge> so maybe it's time to take another approach: spliting the scenario job into 2 jobs (for instance traffic classes vs the rest)
16:13:33 <gthiemonge> what do you think about that?
16:14:10 <tweining> if it makes the execution faster, fine with me
16:14:17 <oschwart> Sounds good to me
16:14:34 <johnsom> We have done that in the past. Though, I have not looked at this patch from froyo to see if we can optimize it better.
16:14:58 <johnsom> I would like to review it before we merge it.
16:15:07 <gthiemonge> running the test_traffic_ops class takes 2h in my env
16:15:09 <gthiemonge> sure
16:15:21 <gthiemonge> we will not merge it until we get acceptable durations for the jobs
16:15:35 <johnsom> plus we probably don't want to merge that at this point in the release as we probably have priority patches we don't want to impact
16:15:48 <gthiemonge> right
16:16:18 <johnsom> repeat=19
16:16:32 <johnsom> Yeah, that is probably more time than we really need to spend in the test
16:17:35 <gthiemonge> in most jobs, this test class finished more than one hour after the other classes (ok we have only 2 concurrent classes at a time)
16:18:56 <gthiemonge> anywys, I will propose a test patch that splits the job, so we can see the durations
16:19:20 <johnsom> Each one of those new test cases will make 19 requests now, so yeah, huge jump in test runtime needed. This needs some review time.
16:19:54 <gthiemonge> note that c9s job fails already with a lot of timeouts
16:20:29 <johnsom> Yeah, agreed, C9S is slow...
16:20:51 <johnsom> I'm just not sure we need to add 228 round trips to the members to have a good test.
16:21:32 <gthiemonge> ok
16:24:20 <gthiemonge> #topic Brief progress reports / bugs needing review
16:25:30 <gthiemonge> on my side, I've worked on these gate issues (pylint/grenade)
16:25:45 <gthiemonge> + reviews
16:25:54 <gthiemonge> + a lot of downstream stuff
16:26:22 <oschwart> Would this failure mean that we still have pylint issues on stable branches?
16:26:25 <johnsom> Lots of review for me
16:26:29 <oschwart> #link https://review.opendev.org/c/openstack/python-octaviaclient/+/869806
16:27:22 <gthiemonge> oschwart: I fixed only the octavia repo, I'll take a look at the octaviaclient, but here it passed 10min ago
16:27:38 <tweining> I updated the patch that adds an tick-tock CI grenade upgrade job
16:27:41 <tweining> https://review.opendev.org/c/openstack/octavia/+/860221
16:28:04 <gthiemonge> oschwart: it seems that we need to backport
16:28:05 <gthiemonge> https://review.opendev.org/c/openstack/python-octaviaclient/+/872357
16:28:57 <johnsom> Oh, nice. The terminology is a mess for tick-tock vs slurp and the docs seem to contradict themselves. I'm glad someone took the time to figure out what needs to happen there.
16:29:05 <oschwart> gthiemonge: I will take care of it if you guys are ok with it
16:29:47 <gthiemonge> oschwart: ok! it seems that it is only needed on zed
16:30:01 <tweining> yeah, I guess SLURP is the correct name now. I can update the patch if you want
16:31:23 <oschwart> gthiemonge: ack
16:31:39 <oschwart> tweining: should we keep the ffu u/s job then?
16:31:43 <tweining> and regarding oschwart's comment about removing octavia-grenade-ffu. I guess he's right. I will update the patch
16:32:22 <tweining> is stein even supported anymore?
16:32:46 <gthiemonge> tweining: haha no
16:33:08 <gthiemonge> tweining: I tried to update the grenade-ffu job, but it was not working at all
16:35:25 <opendevreview> Tom Weininger proposed openstack/octavia master: Add octavia-grenade-slurp CI job  https://review.opendev.org/c/openstack/octavia/+/860221
16:38:56 <gthiemonge> #topic Open Discussion
16:41:57 <opendevreview> Omer Schwartz proposed openstack/python-octaviaclient stable/zed: Add munch explicitly to requirements  https://review.opendev.org/c/openstack/python-octaviaclient/+/873151
16:42:10 <gthiemonge> any other topics folks?
16:43:04 <oschwart> Nothing from me. I will prob ask you guys to review the munch backport patches as we have until next week (?) for the client patches
16:43:22 <gthiemonge> I guess it will be merged before the end of week ;-)
16:44:08 <oschwart> gthiemonge: ack
16:44:13 <gthiemonge> ok folks, thank you!
16:44:16 <gthiemonge> have a great week
16:44:19 <gthiemonge> #endmeeting