16:01:53 #startmeeting openstack-helm 16:01:54 Meeting started Tue Feb 4 16:01:53 2020 UTC and is due to finish in 60 minutes. The chair is lamt. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:55 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:01:57 The meeting name has been set to 'openstack_helm' 16:02:08 o/ 16:02:35 o/ 16:02:36 srwilkers, lamt, portdirect, jayahn, korzen, alanmeadows, mattmceuen, raymaika, evrardjp, MegHeisler, howell, georgk, arunkant, roman_g, dwalt, gagehugo, stevthedev ^ 16:02:37 o/ 16:03:10 O/ 16:03:14 o/ 16:03:25 portdirect is busy so I will be doing hosting duty again - agenda for this week: https://etherpad.openstack.org/p/openstack-helm-meeting-2020-02-04 16:03:43 will give folks a few more minutes 16:05:58 okay let's get rolling 16:06:33 #topic Gate jobs failure 16:07:41 A few weeks ago Pete volunteered to redo the gate with new kubeadm, but there is delay. 16:08:17 and there are a handful of (non-voting) jobs that consistently fail 16:08:46 e.g. openstack-helm-infra-aio-podsecuritypolicy 16:09:02 are they broken because things are being merged in that failed those jobs to begin with? 16:09:20 and/or are they just succumbing to bitrot because they're not being updated accordingly? 16:09:52 I am not sure - if I am to guess, a little bit of both 16:10:23 that's unfortunate 16:10:59 most likely both 16:11:16 we tend to move them to non-voting and/or to experimental, and they kind of get forgotten 16:11:28 we should probably try and carve out some time/effort to get those working again, lest we keep removing jobs just because they're broken 16:11:43 and we just end up burning through open-infra resources for jobs we know and expect to fail 16:12:11 ++ or just remove them if they are not adding value, having a job that fails 100% of the time doesn't make sense imho 16:13:23 well, the jobs we have add value for the most part 16:13:37 i don't think removing them is the true answer 16:13:46 ie, the password rotation jobs 16:13:52 or release uuid update jobs 16:14:25 yeah. We can wait until https://review.opendev.org/#/c/701991/ is done and reevaluate, but it is something I want to raise awareness to those failing non-voting jobs. 16:14:39 but they're only valuable if people actually triage and address their failures 16:14:57 agreed 16:16:57 We can revisit this when that PS gets closer, so we can address all the failing jobs at that point, but that's all I have regarding this topic. 16:18:22 moving on 16:18:35 #topic reviews 16:18:58 https://review.opendev.org/#/c/702983/ - osh-infra bandit 16:18:58 https://review.opendev.org/#/c/705031/ - move apparmor to use feature gates (also makes the job pass now) 16:18:58 https://review.opendev.org/#/c/697554/ - Map LDAP groups in keystone for gate overrides 16:19:40 speaking of making a broken job pass... 16:20:06 there looks to be a lot of incoming apparmor patches 16:20:55 if folks can review them, it would be much appreciated. 16:21:52 thanks for that work, gagehugo 16:22:04 if there's a passing job, i'll review em 16:22:54 ++, most of those apparmor jobs need to be using the feature gate from what I can tell 16:23:12 apparmor patches I should say 16:23:35 that should reduce the number of those bash scripts 16:24:11 yep - that's why that functionality exists 16:25:42 #topic open floor 16:26:09 the agenda is light today - so opening the floor for anyone who has items they wish to discuss 16:28:30 well, if no one has anything, giving everyone 30 minutes back. Have a good rest of the day. 16:28:39 o/ 16:28:44 \o 16:28:57 #endmeeting