[08:15:49] <Emperor>	 Hi folks, could I get a +1 on https://gerrit.wikimedia.org/r/c/operations/puppet/+/1266138 please? we have two rows worth of eqiad systems to reimage, this drains the first 3 swift backends.
[08:20:57] <jynus>	 done
[08:21:49] <Emperor>	 TY
[08:45:37] <jynus>	 reciprocation? https://gerrit.wikimedia.org/r/c/operations/puppet/+/1266148
[08:50:12] <Emperor>	 +1
[08:55:02] <jynus>	 thank you
[10:48:52] <Raine>	 o/, marostegui or Amir1: I pinged on task https://phabricator.wikimedia.org/T419980 about the table swap for the new process for ICU upgrade and didn't get a response, could we please schedule a time to do it?
[10:52:42] <marostegui>	 Hi Raine - I came back from 3 weeks out, so I am still catching up
[10:52:46] <marostegui>	 I will take a look
[10:52:49] <Raine>	 thank you <3
[10:53:42] <Raine>	 if you'd like to have more time before doing the procedure, feel free to suggest a later date, I'd just like to fix one, so CommRel have time to announce it 
[11:13:42] <marostegui>	 Raine: The dates you've proposed will not work, as we have public holidays till tue
[11:14:00] <marostegui>	 And I think we need to discuss the fact that it cannot be done gradually but all or nothing
[11:14:02] <marostegui>	 I will comment on the task
[11:36:56] <Raine>	 ack, thank you marostegui <3 
[11:41:48] <jinxer-wm>	 FIRING: PuppetFailure: Puppet has failed on ms-be1065:9100 - https://puppetboard.wikimedia.org/nodes?status=failed - https://grafana.wikimedia.org/d/yOxVDGvWk/puppet - https://alerts.wikimedia.org/?q=alertname%3DPuppetFailure
[11:48:17] <Emperor>	 4🦊🐟, I've filed one bust-disk ticket today already
[12:09:38] <Emperor>	 T422011 filed for ms-be1065, alert silenced
[12:09:38] <stashbot>	 T422011: Disk (sdt) failed in ms-be1065 - https://phabricator.wikimedia.org/T422011
[12:49:05] <federico3>	 I added some basic docs to https://wikitech.wikimedia.org/wiki/MariaDB#HAProxy
[12:57:07] <marostegui>	 federico3: One clarification, only if the primary goes down it needs restarting, if the backup ones does, and then it comes back, haproxy will pick it up
[13:06:32] <federico3>	 I updated the page to make it more clear
[13:06:57] <marostegui>	 thank you!
[14:01:01] <jynus>	 The reason to not switch back is the lack of stonith, and previous cases where it flapped up and down (like phabricator in the old days with myisam)
[14:01:57] <jynus>	 at some point that architecture should be revised, it was never changed when I onboarded due to vip constraints
[14:02:35] <jynus>	 vip as in floating virtual ip constraints
[18:09:45] <Amir1>	 Another day, another inefficient querying in mw :D https://phabricator.wikimedia.org/T422069