[07:00:36] !log tools delete tools-nfs-2 - T404584 [07:00:41] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [07:00:41] T404584: [tools,nfs,infra] Address tools NFS getting stuck with processes in D state - https://phabricator.wikimedia.org/T404584 [08:47:41] Шаги теперь — это валюта. Пиши — объясню [12:26:29] Does wmcs support nested virtualiation? [12:35:40] !log tools moving toolforge traffic to haproxy directly T283948 [12:35:45] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [12:35:45] T283948: Merge Toolforge Nginx front proxy into the existing K8s HAProxy setup - https://phabricator.wikimedia.org/T283948 [12:40:16] taavi: \o/ \o/ [12:40:46] Damianz: not afaik [12:51:44] Damianz: no, or not yet, we discovered that it prevents live migration T276208 , but might be worth it trying again [12:51:44] T276208: cloud: libvirt doesn't support live migration when using nested KVM - https://phabricator.wikimedia.org/T276208 [12:51:56] (would be very nice also to pre-build lima-kilo) [12:52:07] and run the functional tests with it [12:52:16] (on ci in cloudvps I mean) [13:02:06] exactly [13:02:58] I mean we could do it using just the base image but with puppet etc that would be a lot of moving parts, being able to just spin a vm and run the tests for a PR in what should be a reproducible env would be very nice [15:01:41] Did gitlab get changed recently? It looks kinda funky today e.g. a white outlined logo on a grey background [15:03:27] that's the wikimedia logo I think (or some version of it) [15:05:40] seems like a question for #wikimedia-gitlab [15:12:02] posted there, I'm now just going to slowly walk around blind folded before something else sticks it's head up as needing work [15:29:46] there was a gitlab version upgrade. not aware of specifics regarding the logo but could have influenced something. [15:34:23] Damianz: https://phabricator.wikimedia.org/T407993#11298868 [15:34:39] oh, that's yours, haha. but good :) [15:34:48] no idea who that guy is [15:34:54] lol [15:35:34] That was the reply from #wikimedia-gitlab fwiw [15:35:50] Damianz: will mention it in a private ticket that was about the gitlab upgrade [15:37:05] lol [15:38:12] oh, look, and since security processed it.. it is already public [15:38:35] you will see it in your ticket now. cya [16:24:46] I haven't quite got my head around the bat tests.. but in `creating deploy with same ref reuses the build` am I reading it right that the first build uses the default retry and the second uses 5min? [16:26:04] It's failing with `Retrying 120 times: xxx` because uwsgi is taking more than 2min to build, it's un-clear why 1 specifies 300 and the other nothing. I would sort of expect all build/deployment waiting to be 5min (300s) [18:57:02] Hi all, who do I need to speak to in order to discuss extended capacity for tools on ToolForge? Supporting researchers from King's College who have developed ProVe and need to migrate to ToolForge, but may need more that space than the default. Someone said WMF might extend capacity in certain cases, and we want to explore if we might fall under the right criteria for [18:57:02] that. Thanks in advance! [18:57:38] what kind of space? disk space on the shared NFS? database in ToolsDB? [18:58:13] (I’m not sure we have limits on either in Toolforge… I usually hear about quota increase requests in a Cloud VPS context instead, though I might just be confused ^^) [19:00:49] Lucas, I really don't know. I'm trying to find out who the right person is to talk to the researchers themselves, who know all the technical details. We were in a meeting with WMDE, and they suggested we ask the Cloud folks about it. I understand they somehow want to to manage the tool in two different spaces, but don't have the technical details. They do, hence, asking [19:00:50] about me [19:00:51] eting the right person. :) (re @lucaswerkmeister: what kind of space? disk space on the shared NFS? database in ToolsDB?) [19:01:17] ok, then I’m sure the right person will be around soon enough :) [19:01:22] (it’s not me, I was just curious ^^) [19:10:04] for general questions see https://wikitech.wikimedia.org/wiki/Help:Cloud_Services_communication, or if they already know what exactly they need then https://phabricator.wikimedia.org/project/view/4834/ [19:19:46] Appreciate it, Lucas. :) (re @lucaswerkmeister: ok, then I’m sure the right person will be around soon enough :)) [19:20:51] That link sends here... there's no email, or clear understanding who we need to talk to. They don't have all the answers, so phabricator is not the right course of action in this case.. Thanks, though! (re @wmtelegram_bot: for general questions see https://wikitech.wikimedia.org/wiki/Help:Cloud_Services_communication, or if they already know...) [19:35:13] @Shani: folks can send me emails if they'd like (bd808@wikimedia.org), but really the links Taavi gave are how WMCS does work. [19:41:56] A Phabricator task explaining their needs would be a reasonable way to start a conversation. The main think I see noted at [[d:Wikidata:ProVe]] that Cloud VPS/Toolforge won't have for them is GPUs. Within reason we can support compute and storage needs. [19:42:04] *thing