14:00:33 #startmeeting cinder 14:00:33 Meeting started Wed Nov 29 14:00:33 2023 UTC and is due to finish in 60 minutes. The chair is rosmaita. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:33 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:33 The meeting name has been set to 'cinder' 14:00:40 #topic roll call 14:01:08 howdy 14:01:22 hi 14:01:30 o/ 14:01:50 o/ 14:02:32 o/ 14:02:37 o/ 14:03:11 ok, let's get started 14:03:19 #link https://etherpad.opendev.org/p/cinder-caracal-meetings 14:03:23 meeting agenda ^^ 14:03:42 Rajat is away at the Horizon PTL's wedding today 14:03:46 o/ 14:03:57 nice 14:04:02 so best wishes to Vishal on behalf of the entire cinder team 14:04:17 +1 14:04:19 #topic announcements 14:04:35 first, the midcycle is next week 14:04:38 +1 14:04:40 o/ 14:04:49 it's 2 hours and overlaps with the normal meeting time 14:04:56 o/ 14:05:10 so, to be explicit: 14:05:11 Date: 6th December, 2023 14:05:18 Time: 1400-1600 UTC 14:05:25 Meeting link: https://meet.google.com/auq-jiem-zxy 14:05:32 Etherpad: https://etherpad.opendev.org/p/cinder-caracal-midcycles 14:05:48 we are light on topics, so please add items you want to discuss or have questions about 14:06:08 next up is an issue that came up on the mailing list 14:06:17 #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/VL5QIG4JXOPD3V45MALBMKZFXNNBIWIT/ 14:06:25 asyncore is removed in Python 3.12 14:06:48 it's an issue for us because taskflow uses asyncore, and we use taskflow (a LOT) 14:07:39 you can read through the thread for details, but the short story is that asyncio is supposed to be the replacement for asyncore 14:07:52 but, asyncio does not play nicely with eventlet 14:08:01 and we are pretty big eventlet users 14:08:45 anyway, short-term it is looking like pyasyncore could be used in py 3.12 14:09:16 eventlet also doesn't work yet on py 3.12, so we'll be blocked by that too 14:09:35 that's right, i think we talked about that last week 14:10:13 i think someone is working on getting eventlet to work in py 3.12 14:10:24 but i don't know anything officially about that 14:10:43 looks like we are having some aging infrastructure issues 14:11:24 anyway, if you want to dig into eventlet or taskflow, there are opportunities 14:11:34 i think taskflow is an oslo thing at this point 14:11:59 final announcement is about cinderlib 14:12:15 the Bobcat release has to happen on or before Dec 7 14:13:04 I think the CI is green and there are just a few patches open, nothing that looks critical though 14:13:12 #link https://review.opendev.org/q/project:openstack/cinderlib+status:open 14:13:48 and I think we discussed this at the last video meeting, this will be the final cinderlib release 14:13:58 we won't be continuing development on it 14:14:56 that's all the official announcements, anyone else have something to announce? 14:16:17 guess not 14:16:23 #topic CI Jobs 14:16:38 two jobs I want to talk about 14:16:52 1 - cinder-tempest-plugin-cbak-s3 14:17:05 #link https://zuul.opendev.org/t/openstack/builds?job_name=cinder-tempest-plugin-cbak-s3&project=openstack/cinder-tempest-plugin 14:17:24 looks like it's been failing for a month 14:17:44 and it's blocking the patch to increase the swap space in other jobs 14:17:53 this was the one with the issue that looks related to ssl configuration? 14:18:10 maybe? 14:18:21 cinder.backup.drivers.s3.S3ClientError: S3 client error: An error occurred (400) when calling the HeadBucket operation: Bad Request 14:19:06 i guess this is the more relevant one: 14:19:07 botocore.exceptions.ClientError: An error occurred (400) when calling the HeadBucket operation: Bad Request 14:21:59 anyway, i will look into it unless there's someone here with an interest in the s3 backup driver? 14:22:14 not really familiar with it 14:22:18 so Can't help 14:23:40 2 - grenade 14:23:53 these aren't failing systematically 14:24:24 but it looks like when it does fail, it's happening in the grenade setup 14:24:49 there's a "cinder resource" section where a server gets created and then 3 volumes get created and attached 14:25:11 sometimes the third volume doesn't get to in-use before the grenade script times out 14:25:40 other times, the third volume goes to in-use, but then the next step is to ssh into the server, and that fails for some reason 14:26:34 we need to start tracking these 14:27:03 :-( 14:27:54 #link https://etherpad.opendev.org/p/cinder-caracal-ci-tracking 14:28:32 so, if you see a grenade failure, please put a link to the logs in ^^ 14:28:51 it would be even better if you could say what the failure is 14:29:32 i'll look for an example of each of the failures i mentioned earlier and put it on the etherpad so you know what to look for 14:30:24 thks 14:30:32 ok, that's all from me 14:30:44 #topic image metadata values 14:31:10 drencrom: this is the last item, maybe we can end the meeting early and work on the etherpad 14:31:23 sure 14:31:52 do you have a link handy? 14:32:12 https://etherpad.opendev.org/p/cinder-glance-nova-image-metadata-value-length 14:32:24 thanks! i have too many windows open 14:32:54 ok, that's all the official items for today 14:33:27 i propose that we end the meeting early, and me and drencrom work on the image metadata email, and everyone else take a half-hour to review the requests on the etherpad 14:34:09 hearing no objections, i declare the meeting over! 14:34:12 #endmeeting