15:00:26 <gagehugo> #startmeeting openstack-helm 15:00:28 <openstack> Meeting started Tue May 11 15:00:26 2021 UTC and is due to finish in 60 minutes. The chair is gagehugo. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:29 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:31 <openstack> The meeting name has been set to 'openstack_helm' 15:00:43 <stevthedev> Hello there, general ghugo 15:01:15 <lamt> \o 15:01:20 <sangeet> 0/ 15:01:26 <jinyuanliu> \o 15:01:27 <gagehugo> #link https://etherpad.opendev.org/p/openstack-helm-weekly-meeting agenda 15:01:48 <reddy1> \o 15:02:03 <gagehugo> we will start in a few 15:02:49 <miniroy> o// 15:04:49 <gagehugo> #topic OSH Announcement 15:05:35 <gagehugo> I'd like to announce that we are adding sangeet and jinyuanliu to the OSH core team, congratulations to you both 15:05:46 <stevthedev> \o/ 15:05:50 <miniroy> woot woot 15:06:14 <sangeet> Thank you 15:06:23 <jinyuanliu> thank you 15:07:02 <gagehugo> thank you both for all the work you've done, we hope to see it continue :) 15:07:53 <gagehugo> #topic pod localtime 15:08:05 <gagehugo> jinyuanliu: I think this is your topic 15:08:13 <jinyuanliu> yes 15:08:21 <jinyuanliu> https://review.opendev.org/c/openstack/openstack-helm/+/789135 15:09:10 <jinyuanliu> I believe that more places will change the time zone according to their geographic location. 15:09:10 <jinyuanliu> For example, in China, 15:09:10 <jinyuanliu> we generally use CST. I believe that other regions have similar problems, which will cause inconsistencies in the container with the host time, 15:09:10 <jinyuanliu> and some problems will occur.Especially the time of the log is wrong, which is not conducive to viewing。 15:09:28 <jinyuanliu> It seems that there are currently several ways to set the POD time. 15:10:07 <jinyuanliu> TZ can be set by env, but this requires the installation of tzdata pkg. 15:10:08 <jinyuanliu> I have tested it and it is invalid to set TZ without this package. 15:10:18 <jinyuanliu> We can install pkg of the images through opensatck-images, but some of the images we directly pull from GitHub seem to be difficult to handle. 15:10:19 <jinyuanliu> RabbitMQ, for example, does not have tzdata installed. 15:10:37 <jinyuanliu> and 15:10:39 <jinyuanliu> PodPreset is also a way, as samuei said,PodPreset(alpha feature) has been deprecated , see: https://github.com/kubernetes/kubernetes/pull/94090 15:10:58 <jinyuanliu> MountPath seems to be the best 15:11:58 <jinyuanliu> What do you think?Is this necessary? 15:12:47 <sangeet> I am sure all host should have the same time, I do not see any issues with mounting host time to the pods 15:12:54 <gagehugo> Won't this break anyone currently using UTC? 15:13:32 <sangeet> If the hosts have UTC then the pods will have UTC as well when mounting the host time 15:14:10 <miniroy> I am more worried about things like centralized logging or where some place may have some aggregate monitoring setup for different geographic regions 15:14:26 <lamt> can we make that optional? 15:14:27 <miniroy> for a localize datacenter, this maybe fine 15:14:38 <lamt> instead of always mounting it 15:14:43 <gagehugo> I'd be ok if this was toggle-able 15:14:44 <miniroy> yeah, I think make it configurable and optional maybe better 15:14:50 <lamt> if someone wants they can use it 15:14:52 <sangeet> I am good with making it optional 15:15:02 <lamt> but we won't forcefully mount it in the chart 15:15:17 <sangeet> what do we want the default behaviour to be? 15:15:30 <gagehugo> default should be host time IMO 15:15:42 <gagehugo> how it is now 15:15:58 <sangeet> I agree 15:16:16 <gagehugo> then add a local-time or local_time config option in values.yaml 15:16:39 <gagehugo> I think that would solve most issues if anyone wants to use local time over UTC 15:16:49 <lamt> yeah let's leave the default as is and if you want to test it - use the override 15:17:00 <miniroy> +2 15:17:20 <gagehugo> keystone tokens may break too across geographic regions if we change all the times as well 15:17:37 <jinyuanliu> I agree 15:19:37 <gagehugo> sounds good 15:20:00 <gagehugo> jinyuanliu: you have a path forward then? 15:20:43 <jinyuanliu> Yes i will update a path 15:21:27 <gagehugo> cool 15:21:33 <gagehugo> #topic open discussion/reviews 15:21:45 <gagehugo> Anyone have anything they want to discuss or anything to get reviewed? 15:23:07 <gagehugo> #link https://review.opendev.org/c/openstack/openstack-helm/+/789948 15:23:14 <gagehugo> #link https://review.opendev.org/c/openstack/openstack-helm/+/789949 15:23:18 <gagehugo> I'll post some then :D 15:23:48 <sangeet> https://review.opendev.org/c/openstack/openstack-helm-infra/+/788061 15:24:33 <sangeet> lamt .. renewBefore field is set by default in all regular certificates but not in ingress certificates. So may not be a good idea on relying on this field. 15:24:58 <lamt> I need to double check - I still don't think deleting the secret is needed 15:25:11 <miniroy> with the addition of jinyuanliu and sangeet, we have finally achieved "follow the sun" support :) 15:26:18 <sangeet> yes .. the certs are rotated after renewalTime and secrets updated, but how do we mount new secrets 15:27:17 <lamt> doesn't restarting the pod work with the new secrets? 15:27:41 <sangeet> And this is a cron job, the certs may have just rotated and cron would exit without doing anything 15:28:02 <sangeet> I mean certs roatetd before the cron started 15:29:32 <lamt> I haven't touch cert-manager for a while so I need to spend time double checking 15:30:55 <lamt> but having a cron job that deletes things independent of the cert-manager's management of certs/secrets makes me slightly uncomfortable, but if the other cores are okay with it 15:36:53 <gagehugo> Thanks everyone, have a good rest of the week 15:36:55 <gagehugo> #endmeeting