[07:38:35] Good morning! [08:28:07] I resent the invitations for this week's team meetings [08:41:47] Morning :) [08:42:12] I'll be out for ~2h after ten, for an appointment, but back around lunchtime [08:46:06] (03CR) 10Ilias Sarantopoulos: Makefile: add support for article-descriptions (032 comments) [machinelearning/liftwing/inference-services] - 10https://gerrit.wikimedia.org/r/993481 (https://phabricator.wikimedia.org/T356176) (owner: 10Kevin Bazira) [08:50:29] Morning Tobias! [08:50:34] morning o/ [08:50:42] Hey Aiko! [08:51:09] hi Ilias :) [08:51:32] I wanted to sync about multilingual and kserve. I'm going afk for an errand but I'll be back in like 30' [08:57:24] (03CR) 10Kevin Bazira: Makefile: add support for article-descriptions (032 comments) [machinelearning/liftwing/inference-services] - 10https://gerrit.wikimedia.org/r/993481 (https://phabricator.wikimedia.org/T356176) (owner: 10Kevin Bazira) [09:32:09] isaranto: ok o/ [10:08:53] o/ [10:09:14] aiko: do you want to have a call now instead of later? [10:17:11] isaranto: yeah we can do that :) [10:20:09] isaranto: in 10min? [10:21:03] aiko: yep! [11:50:31] heads up: I am going to drain the staging workers in turn to update everything to use the newest version of runc (as discussed with Moritz yesterday). [11:52:04] ack! [11:55:10] * isaranto afk lunch! [12:05:08] Ok, all done, and pods are running again [12:05:52] moritzm: all pods in staging have been restarted. I recommed we let this soak 'til tomorrow and do the serving cluster in eqiad then [12:06:55] sounds good [12:07:28] while we're at it, are ther kernelupdates/rebooting-would-help things pending? Might as well doa reboot too for that if we're draining hosts [12:13:54] (03CR) 10AikoChou: "I agree to proceed with kserve 0.11.2" [machinelearning/liftwing/inference-services] - 10https://gerrit.wikimedia.org/r/995198 (https://phabricator.wikimedia.org/T356501) (owner: 10AikoChou) [12:28:03] (03PS5) 10AikoChou: Makefile: add support for revertrisk-multilingual [machinelearning/liftwing/inference-services] - 10https://gerrit.wikimedia.org/r/995198 (https://phabricator.wikimedia.org/T356501) [12:55:26] (03CR) 10Kevin Bazira: Makefile: add support for revertrisk-multilingual (031 comment) [machinelearning/liftwing/inference-services] - 10https://gerrit.wikimedia.org/r/995198 (https://phabricator.wikimedia.org/T356501) (owner: 10AikoChou) [15:33:05] 10Machine-Learning-Team, 10Patch-For-Review: Support running revertrisk-multilingual model-server via Makefile - https://phabricator.wikimedia.org/T356501 (10isarantopoulos) [15:58:55] 10Machine-Learning-Team, 10SRE, 10Patch-For-Review: Requesting write access to ml-staging-codfw for ML team - https://phabricator.wikimedia.org/T354516 (10isarantopoulos) I tried to delete a revision and an inferenceservice on experimental namespace and it seems that I don't have access: ` kubectl delete re... [16:02:09] klausman: --^ I update the task. I also tried to patch an inference service with the same results [16:02:34] ty! [16:06:16] for now I removed the gpu manually with a patch , cause I want to run a load test with the cpu only https://gerrit.wikimedia.org/r/c/operations/deployment-charts/+/997901 [16:07:39] Yeah, I saw it (re) scheudle when doing the staging drains and got excited for a moment :D [16:10:07] kevinbazira: manually downloading the model worked \o/ [16:11:29] (03PS6) 10Kevin Bazira: Makefile: add support for article-descriptions [machinelearning/liftwing/inference-services] - 10https://gerrit.wikimedia.org/r/993481 (https://phabricator.wikimedia.org/T356176) [16:11:39] so if you can add the `--continue` option to wget and update the commit msg we'll be ready to go [16:12:18] nice :D already done [16:12:56] (03CR) 10Ilias Sarantopoulos: [C: 03+1] Makefile: add support for article-descriptions [machinelearning/liftwing/inference-services] - 10https://gerrit.wikimedia.org/r/993481 (https://phabricator.wikimedia.org/T356176) (owner: 10Kevin Bazira) [16:13:27] isaranto: good to know it has finally worked :D [16:13:37] nice work! [16:14:20] thanks to all of you for the invaluable input in the meeting :) [16:14:29] <3 [16:26:34] (03CR) 10AikoChou: [C: 03+1] Makefile: add support for article-descriptions [machinelearning/liftwing/inference-services] - 10https://gerrit.wikimedia.org/r/993481 (https://phabricator.wikimedia.org/T356176) (owner: 10Kevin Bazira) [16:31:56] (03CR) 10Kevin Bazira: [V: 03+2 C: 03+2] "Thanks for the reviews :)" [machinelearning/liftwing/inference-services] - 10https://gerrit.wikimedia.org/r/993481 (https://phabricator.wikimedia.org/T356176) (owner: 10Kevin Bazira) [17:43:05] isaranto: one thing we could try tomorrow is wothout the v1 suffix. Maybe Alex was wrong about those. [17:43:58] that's the only thing I could come up with, especially since all the other roles doent have version suffixes specified [17:44:37] My gut tells me that maybe versions are either not part of the schema at that point at all, or have a different syntax [17:44:54] anyway, heading out for now. Seeya tomorrow! [17:45:24] Oh and re: Analytics downloads, I'm coordinating with Ben over in #wikimedia-analytics [17:51:04] ciao Tobias, I'm heading out as well o/