[00:05:23] * bd808 off [06:14:03] andrewbogott: i figured I'd do most / al of the work :-) [07:15:02] morning [07:16:05] o/ [07:46:07] o/ [07:57:51] morning [08:00:27] o/ [08:00:52] I would be interested in presenting something regarding Toolforge at Kubecon, but the EU version, not NA :-P [08:11:01] dcaro: could you take a look at this to see if it's correct and/or if it's missing something? T362299 [08:11:02] T362299: [api-gateway] Add a python server to serve consolidated openapi docs - https://phabricator.wikimedia.org/T362299 [08:14:24] 👀 [08:29:00] blancadesal: added a couple comments, looks good [08:56:02] dcaro: thanks! [09:03:59] dcaro: about the oapi-server, what I meant was to skip tox/pre-commit etc. altogether in the very first iteration: just get the basic plumbing together to serve hello world from the server, then add business logic + CI stuff after that [09:05:15] blancadesal: I would start with the ci, and then add the plumbing xd, but I guess it's up to you [09:07:06] I like to start with the part that I feel most unsure about xd [09:07:06] which in this case is the helm/k8s stuff [09:09:08] do you think starting with the ci would be easier? [09:19:16] I guess that for me yes, that way I don't have to do any linting/test fixes/code setup changes/etc. once I introduce the code, and can go testing the generated images right away, but I suppose it's a personal choice [09:35:37] can I get +1 on https://phabricator.wikimedia.org/T361895 ? [09:36:09] dcaro: LGTM [09:36:12] dcaro: done [09:41:02] thanks! [10:17:26] I'm playing caching the pre-commit setup to speed up CI with https://gitlab.wikimedia.org/repos/cloud/cicd/gitlab-ci/-/merge_requests/31, it's been misbehaving a lot lately (ci runs timing out at 1h... my guess is that we got blocked somehow at github level) [10:48:45] dcaro: +1'd [10:51:18] thanks, I'm thinking I'll sync the versions first and avoid creating a 2G image [14:41:18] dcaro: the ceph cookbooks need to be migrated from wmcs_libs.alerts to spicerack.alertmanager, do you want to do that as you're already working on those cookbooks or should I send patches for you to review? [14:42:02] you can leave it to me if you want, I have to do another round on them, can you add a comment in the patch so I don't forget though? [14:42:04] (any will do) [14:44:51] updates for everyone! https://gitlab.wikimedia.org/groups/repos/cloud/-/merge_requests?scope=all&state=opened&label_name[]=Needs%20review&approved_by_usernames[]=None [14:44:52] xd [14:45:06] first successful run of a scheduled pipeline to try to update pre-commit hooks [14:45:35] should the '# do not update this until our CI runner uses Python >=3.8.1' comments be removed now? [14:45:52] yep I think so [14:46:17] I'll go manually through the ones open and fix them (some are failing because of an old includes in the gitlab yaml I think) [14:46:57] it's using my personal token as I was not able to find a way yet to use the group one (it might not be available :/) [14:48:51] feel free to review https://gitlab.wikimedia.org/repos/cloud/cicd/gitlab-ci/-/merge_requests/32 :), that's the pipeline, I'll remove the bits to make it work unmerged before merging once approved [14:53:43] anyone wants to review the pre-commit autoupdate MRs while I fix them? Or can I just merge them once they pass the tests? [14:56:15] i'm fine with you just self-merging, but can also review things if you prefer that way [15:51:03] * arturo offline [15:56:11] taavi: I've removed the 'needs review' from the ones I need to change, all the ones left with the flag are ready to get reviews :) [15:56:18] taavi: can you advise about the last comment on T361772? At this point I'm reluctant to try to fix it without supervision :) [15:56:21] T361772: Expired cert failure on cloudinfra-cloudvps-puppetserver-1.cloudinfra.eqiad1.wikimedia.cloud - https://phabricator.wikimedia.org/T361772 [16:01:56] andrewbogott: `strace`ing the process reveals /srv/puppet/server/ssl/certs/ca.pem is the expired copy of the CA cert. so backing that up and then replacing it with the renewed version should fix it [16:02:11] hm, so that's basically the same issue that all the clients had [16:02:15] * andrewbogott tries it [16:02:25] * andrewbogott thinks he already tried that yesterday but tries it again [16:04:34] ok, seems to have worked, thanks [18:17:57] * bd808 lunch [23:56:08] * bd808 off