[07:21:58] <elukey>	 hello on-callers
[07:22:14] <elukey>	 in a bit I am planning to start the reimage of one chartmuseum host
[07:22:27] <elukey>	 I'll depool it first
[07:22:50] <elukey>	 they are used as "storage" (via swift) for helm/helmfile deployments
[07:30:33] <elukey>	 I am also going to move the thumbor's eqiad pods to poolcounter1006, the new bookworm VM (already done for codfw, this is to complete the work
[07:44:30] <elukey>	 thumbor done, officially the first one completely on new poolcounter nodes
[07:44:37] <elukey>	 proceeding with chartmuseum2001
[08:29:13] <hnowlan>	 thanks elukey 
[08:35:52] <vgutierrez>	 hmm ldap.toolforge.org is currenly down (returning 502s)
[09:36:40] <vgutierrez>	 dcaro: ^^ if you have a spare minute could you check it? 
[09:41:16] <dcaro>	 vgutierrez: let me have a look yep
[09:41:23] <vgutierrez>	 thx <3
[09:45:42] <elukey>	 chartmuseum2001 up and running
[09:45:45] <elukey>	 going to repool it in a bit
[09:45:50] <elukey>	 (on bookworm)
[09:47:33] <dcaro>	 vgutierrez: restarted it and it's back up, I'm not sure what was the issue though, will try to take a look
[13:45:03] <elukey>	 bd808: o/ when you have a moment could you check https://gerrit.wikimedia.org/r/c/wikimedia/toolhub/+/1071804 ? Nothing super urgent but I'd need a rebuild to pick up git upgrades, and CI seems not linking it :) 
[13:59:49] <swfrench-wmf>	 FYI, e.ffie and I will be running the `sre.switchdc.mediawiki` cookbook in `--live-test` mode for testing ahead of the switchover [0] in a little bit (10-15m or so).
[13:59:49] <swfrench-wmf>	 unlike `--dry-run`, this will be performing actual operations, but is not expected to cause any problems (e.g., is a switch "toward" the current primary DC, and skips disruptive operations).[0] https://wikitech.wikimedia.org/wiki/Switch_Datacenter#Live_Test
[14:12:25] <effie>	 cc  arnaudb, eoghan ^ :)
[14:12:34] <eoghan>	 rgr
[14:14:33] <arnaudb>	 thanks for the heads up :D
[14:35:32] <swfrench-wmf>	 FYI, we're starting this now
[14:37:19] <volans>	 o/
[14:51:38] <swfrench-wmf>	 arnaudb: FYI, during part of this test, the primary DB hosts will temporarily have read-only alert downtimes in place (now present). I expect those to be gone in the next 10m or so.
[14:52:22] <arnaudb>	 ack, thanks for letting me know
[15:03:50] <arnaudb>	 swfrench-wmf: could you let me know when I can move stuff around again? I don't want to cause any parasitic noise in your test :D
[15:04:47] <swfrench-wmf>	 arnaudb: we're on the very last step now, which will also remove the downtimes I mentioned - ETA another couple of minutes
[15:14:16] <swfrench-wmf>	 arnaudb: all done :)
[15:16:15] <arnaudb>	 thanks!
[22:19:37] <bd808>	 elukey: django-elasticsearch-dsl has apparently made a bad https://pypi.org/project/django-elasticsearch-dsl/7.2.2/ release that includes a broken dependency declaration "elasticsearch-dsl (>=7.2.0<8.0.0)". Fixed by https://github.com/django-es/django-elasticsearch-dsl/commit/0f67504927c989e77f7ff39d418db7a798eece15 in their 7.3 release.
[22:26:31] <bd808>	 What is maybe weirder is that we have had django-elasticsearch-dsl==7.2.2 as the pinned dependency since https://gerrit.wikimedia.org/r/c/wikimedia/toolhub/+/866805. I'm not sure what this means changed in that environment. We should be using the same python version (modulo security backports), the same poetry version (again modulo security backports), and the same django-elasticsearch-dsl library version.