[02:57:37] !log tstarling@tools-bastion-15 tools.zoomviewer Stopped the webservice [02:57:39] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.zoomviewer/SAL [10:25:57] !log codesearch hard reboot codesearch9 [10:26:00] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Codesearch/SAL [11:01:17] !log tools starting Kyverno upgrade T427928 [11:01:22] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [11:01:22] T427928: Upgrade Kyverno to 1.16 series - https://phabricator.wikimedia.org/T427928 [15:32:23] !log cluebotng-editsets Added damian as project member (T428432) [15:32:26] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Cluebotng-editsets/SAL [15:32:27] T428432: Horizon project access to cluebotng-trainer adn cluebotng-editsets needed for DamianZaremba - https://phabricator.wikimedia.org/T428432 [15:33:18] !log cluebotng-trainer Added damian as project member (T428432) [15:33:20] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Cluebotng-trainer/SAL [16:16:28] does anyone know who built deployment-wikikube-v127? It's in deployment-prep so the list of candidates is very long [16:35:33] bd808: ^ [16:36:50] if it's him then he already got the email :) [16:39:54] andrewbogott: the Beta Cluster magnum cluster is my work, yeah. [16:40:23] ok! for a moment I thought there might be a fifth phantom magnum user [16:40:41] I was surprised to see Fedora [16:41:11] the upstream service picks its own base images [16:41:12] They're black-box managed worker nodes, doesn't necessarily matter what distro. [16:41:55] Upstream actually /builds/ debian workers now, but there's a bug with debian and I got tired of waiting for it to get fixed/acknowledged. [16:42:29] https://gitlab.wikimedia.org/cloudvps-repos/deployment-prep/tofu-provisioning/-/blob/main/modules/kubernetes/main.tf is what built the deployment-wikikube-v127 cluster. [16:43:28] * andrewbogott cheers for source control [16:43:50] gitops is fun when it works [16:45:01] aha, interesting [16:45:45] The issue with debian workers is silly: https://github.com/vexxhost/magnum-cluster-api/issues/1057 [16:46:16] I considered just standardizing on debian anyway but then thought that bd808 wouldn't appreciate being told "first, make a template that's wrong and then edit it to be right" as the only valid workflow. [16:46:47] I might find that to be a implementation detail best handled upstream, yes. ;) [16:47:04] If the code wasn't in rust I'd've fixed it myself already. [16:47:24] Having debian base images that also work with Puppet might be worth the pain to me though... [16:47:49] I haven't really thought about puppet. Cumin, definitely. [16:48:08] To get puppet working properly we'd have to build our own images; I don't know if that's hard or easy. [16:48:37] andrewbogott: I see claude has already vibe coded its way into that repo, so I guess if you aren't morally opposed you could sling a clanker slop patch at them. [16:48:57] (the capi system is a matryoshka set of black boxes so I don't want to get into hooking the worker orchestration) [16:49:40] I'm a little bit morally opposed but mostly I've made it this long without running claude /at all/ so might as well postpone the learning curve as long as possible. [16:58:33] :empathy: