16:03:45 #startmeeting openstack_ansible_meeting 16:03:46 Meeting started Tue Jun 30 16:03:45 2020 UTC and is due to finish in 60 minutes. The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:03:48 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:03:50 The meeting name has been set to 'openstack_ansible_meeting' 16:03:52 #topic office hours 16:03:56 \o/ 16:04:39 Jonathan Rosser proposed openstack/openstack-ansible-plugins master: WIP add lxd support https://review.opendev.org/738657 16:05:40 so 16:05:51 have we backported everything we wanted regarding focal? 16:07:12 jrosser: btw regarding this... I was thinking if we should just rename containers in dynamic_inventory? 16:07:43 o/ hello 16:08:06 oh you mean _ vs - ? 16:08:13 yeah 16:08:27 oh yes there is probably something nice we need to do there 16:08:44 i sort of put it to one side because theres a bigger piece of work to do with the group names too 16:10:24 i did a reasonable look across the master/ussuri branches for stuff to backport for focal 16:10:29 it looked in reasonable shape 16:10:53 I also had a brief look and didn't find anything at glance 16:11:03 ok, cool, so will place a bump today 16:11:22 a nasty thing is that train bump seems broken... 16:11:50 Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible stable/train: Bump SHAs for stable/train https://review.opendev.org/735469 16:12:58 maybe it's because of some ceph-ansible update... 16:15:56 yes i was looking at that a bit 16:16:10 and i was also thinking that there was something not right with radosgw authentication 16:16:20 I thougnth it might be swift tempest plugin, but it seems that's a part of tempest core. 16:16:26 there is ceph 14.2.10 i think 16:16:36 thats a very recent release which is what could have broken 16:17:33 theres an error in the tempest log about 404 instead of 401, then you can see the auth trail looking all reasonable through radosgw logs and keystone log 16:17:49 and then i didnt really know what was breaking 16:18:34 Like interesting part that tempest was expecting unathorized anyway 16:19:16 but got 404 instead. 16:19:51 so this thing wasn't expected to finish with 200 or smth 16:20:57 so yeah, probably ceph has changed smth in terms of return codes and now it's not fully compatible with swift in terms of return codes... 16:23:44 there was an interesting string in one of the log messages - don't have that open right now 16:24:08 and it took me to a RH bug about ResellerAdmin needing to be in the radosgw config 16:26:35 so here where it fails https://opendev.org/openstack/tempest/src/branch/master/tempest/scenario/test_object_storage_basic_ops.py#L65 16:27:47 so we're checking that things fail here eventually 16:28:12 Jonathan Rosser proposed openstack/openstack-ansible master: WIP - lxd containers https://review.opendev.org/738599 16:29:54 and that code is not touched in several years 16:30:15 i don't know if it's possible to pin ceph back to the previous point release and see what happens 16:30:18 yeah, so I think it's ceph who just returns 404 instead of 401 for unathorized 16:31:14 btw I guess that's not happening for O 16:32:18 Jonathan Rosser proposed openstack/openstack-ansible master: WIP - lxd containers https://review.opendev.org/738599 16:32:41 right, which is wierd, as you'd expect all the changes on N to be backports from O 16:38:38 I don't see anything obvious here https://github.com/ceph/ceph/compare/v14.2.9...v14.2.10 16:42:19 also interesting if old bump is going to pass now https://review.opendev.org/#/c/735469/ 16:52:21 Merged openstack/openstack-ansible-os_rally master: Do not install development packages on the target https://review.opendev.org/731459 16:52:38 Jonathan Rosser proposed openstack/openstack-ansible-os_rally stable/ussuri: Do not install development packages on the target https://review.opendev.org/738676 17:06:01 #endmeeting