[06:55:19] !log tools.versions Udpate to f583b56 [06:55:22] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.versions/SAL [08:23:42] !log tools.ifttt-dev Update to aad850d (T294448) [08:23:45] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.ifttt-dev/SAL [08:56:48] !log tools.ifttt-dev Update to a4c8e30 (T294448) [08:56:52] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.ifttt-dev/SAL [09:38:25] bd808: here’s your patch :P I also created a task so it can be on the projects board https://gerrit.wikimedia.org/r/c/operations/software/tools-webservice/+/921620 [09:39:28] since you (iirc) had mentioned k8s probes in one of the sessions yesterday [09:40:21] I'd love to have probes support in toolforge-jobs and webservice [10:06:17] re probes, once the logging framework exists I think it'd be great to have a standard health endpoint URL that the logging framework filters out from the access logs by default [10:06:52] that's currently the only reason I use a startup probe instead of readiness or liveness, I don't want a ton of /healthz in uwsgi.log :D [10:16:12] you could do something like https://stackoverflow.com/questions/25679166/uwsgi-disable-logs-for-given-urls#25686557 [10:57:07] eh, for an individual tool I’m fine with startup probes only [10:57:20] I just think the logging framework is an opportunity to solve this in a central place [10:57:24] but it’s not urgent or anything ^^ [11:08:12] taavi: any idea how to correctly escape "toolforge.org/started-at" for the field selector? [11:08:18] I’m trying all sorts of variations of [11:08:19] kubectl get pods --field-selector 'metadata.annotations.toolforge\.org/started-at=2023-05-21T11:06:57.174016' [11:08:22] and none of them work [11:08:27] (tool-lucaswerkmeister-test namespace) [11:10:06] did I create a ticket for the liveness probes thingy already? [11:10:26] I did, yes: T335592 [11:10:26] T335592: Toolforge jobs: consider having a way for jobs to report their liveness status to kubernetes - https://phabricator.wikimedia.org/T335592 [11:10:45] lucaswerkmeister: looks like you can't per https://github.com/kubernetes/kubernetes/issues/18801 [11:10:53] arturo: there is a separate one for webservice somewhere too [11:11:09] taavi: ok, let's mention them in each other? [11:12:28] taavi: thanks, I’ll mention that on Gerrit then [11:12:45] although I'm not finding that task [11:14:46] btw here’s what the rolling restart looks like in a kubectl get pods loop (putting the image here since it’s easier than on gerrit ^^) : https://tools-static.wmflabs.org/bridgebot/d841d7e8/file_48365.jpg [11:27:24] dhinus: https://phabricator.wikimedia.org/T337190 [11:29:04] !log tools.bodh bump +1 service resource quota (T337190) [11:29:07] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.bodh/SAL [11:29:08] T337190: Request increased quota for bodh Toolforge tool - https://phabricator.wikimedia.org/T337190 [11:31:52] dhinus: https://netbox.wikimedia.org/ipam/ip-addresses/4792/ [17:49:58] How can I see my contribs via gerrit or get to the global language files? [17:50:08] for WMF production [18:00:01] What do you mean by global language files? [18:13:19] the interface messages, sorry [18:14:46] like the ones that are used to then translate into each language [18:21:42] found it