[07:23:57] FIRING: SystemdUnitFailed: pt-heartbeat-wikimedia.service on db1268:9100 - https://wikitech.wikimedia.org/wiki/Monitoring/check_systemd_state - https://grafana.wikimedia.org/d/g-AaZRFWk/systemd-status - https://alerts.wikimedia.org/?q=alertname%3DSystemdUnitFailed [07:28:50] ^expected [08:18:57] RESOLVED: SystemdUnitFailed: pt-heartbeat-wikimedia.service on db1268:9100 - https://wikitech.wikimedia.org/wiki/Monitoring/check_systemd_state - https://grafana.wikimedia.org/d/g-AaZRFWk/systemd-status - https://alerts.wikimedia.org/?q=alertname%3DSystemdUnitFailed [08:20:12] being setup? [08:20:58] yes [08:21:32] Let me know if you need help, e.g. provisioning them from dbprovs [09:51:03] I got now the codfw warning: Last snapshot for x1 at codfw (db2197) taken on 2026-08-05 23:00:06 is 386 GiB, but the previous one was 434 GiB, a change of -11.0 % [10:20:06] marostegui: the ops.wmf_slave_wikiuser_sleep event ... it seems as though it can produce errors (unkown thread_id when attempting to kill by the look of it) in the logs, is that common? [10:23:31] yeah, that typically happens when there's a race condition on attempting to kill a connection that is already gone. That's common [10:29:09] Wouldn't it make sense to handle that? [10:30:05] I don't think it is causing lots of issues [10:30:38] We can create a task with low priority if you'd like, but I don't think it is something we have to address quickly [10:32:27] Sure, it runs every 30s anyway and the gaps in the logs are much longer than that. I was just thinking about if there are queries that should be killed after that disconnected one causes the abort. [10:33:21] I'll add it in status "refine" for now [10:33:36] sure, and make sure to state it as low priority [10:33:53] and that one kills only sleep ones anyway [14:30:14] https://phabricator.wikimedia.org/P95935 <-- anyone care to tell me how to make "swift stat" tell me about this object? [14:33:15] Ah, got it [14:33:30] printf -v x "1/11/Flag_of_Sri_Lanka.svg/22px-\r\n Flag_of_Sri_Lanka.svg.png" [14:33:43] swift stat wikipedia-commons-local-thumb.11 "$x" [14:34:06] (there are 6 spaces between \n and F) [14:34:40] I'm pretty sure we should delete that, and not just because it's annoying and surely impossible to actually have in a wiki page... [14:34:54] also it's from 2017, so I think escaped Amir1's deletion of thumbs from that container [16:19:01] I think my script doesn't understand files with \n in them :( [16:19:58] everything else is fine