[06:50:40] <ryankemper>	 Gonna take friday off as PTO. I’ll still be around the house for the most part so feel free to ping if I’m needed
[07:13:27] <gehel>	 ryankemper: enjoy the day off!
[07:35:57] <dcausse>	 gehel: if you have a sec: https://gerrit.wikimedia.org/r/c/operations/alerts/+/1035634 (this alert got deployed yesterday but it's getting quite noisy)
[07:36:19] <gehel>	 looking
[07:37:01] <dcausse>	 thanks!
[07:38:00] <gehel>	 merged
[08:31:54] <gehel>	 weekly status update published: https://wikitech.wikimedia.org/wiki/Search_Platform/Weekly_Updates/2024-05-24
[09:53:36] <dcausse>	 lunch+errand
[12:54:44] <inflatador>	 <o/
[13:18:47] <inflatador>	 created a task for CPU performance testing in T365814 ...was thinking a backfill test from cloudelastic might be a good test? Open to suggestions
[13:18:48] <stashbot>	 T365814: Test whether or not CPU performance governor helps Elasticsearch performance - https://phabricator.wikimedia.org/T365814
[13:29:30] <dcausse>	 o/
[14:02:19] <gehel>	 dcausse: steering committee in https://meet.google.com/apj-mcuy-ecn
[15:01:22] <dcausse>	 going offline, have a nice week-end
[15:04:14] <ebernhardson>	 \o
[15:04:22] <dr0ptp4kt>	 peace! i'm gonna drive the vehicle for an hour so sarah can take a call. then i'll switch back to the passenger seat
[15:04:45] <dr0ptp4kt>	 have a good weekend gehel and dcausse 
[15:05:03] <dr0ptp4kt>	 thanks for the edification on the various streaming updater stuff today, also - very much appreciated
[15:05:36] <ebernhardson>	 looks like backfilling is working.  Of course i started this with and older version, and rewrote 1/3 of the thing (but not the general shape) since i started it so it's not really representative
[15:05:41] * dr0ptp4kt sees https://docs.aws.amazon.com/neptune/latest/userguide/sparql-query-hints.html
[15:06:38] <dr0ptp4kt>	 \o/
[15:09:01] <ebernhardson>	 I kinda went around normal code review for this,  and it's obviously impossible to offer truly useful review on a mostly completed thing, but if anyone has ideas: https://gitlab.wikimedia.org/repos/search-platform/cirrus-reindex-orchestrator
[15:09:32] <ebernhardson>	 CI doesn't work yet :P but it runs locally. will get it going
[15:59:51] <inflatador>	 workout, back in ~40
[16:17:14] * ebernhardson ponders a not-terrible name for the (dbname, replica) pairs. Ex: (zhwiki, eqiad), (aawiki, cloudelastic), etc.
[16:48:55] <inflatador>	 back
[17:02:47] * ebernhardson called it CirrusReplicaForWiki. Which is terrible and will have to change :P
[17:40:52] <inflatador>	 lunch, back in ~40
[18:45:40] <inflatador>	 sorry, been back
[19:07:39] <inflatador>	 c-danis posted a jaeger trace in #sre . No problem or anything, just interesting. I noticed wdqs-internal features prominently https://trace.wikimedia.org/trace/2ba1f2755f6f3d4ebd35cb95d1430197
[19:10:30] <ebernhardson>	 nifty. looks like wikidata edits
[19:11:42] <ebernhardson>	 at least, my guess is it would be constraint checking during edits
[19:12:25] <inflatador>	 Does wikidata use internal exclusively? I was wondering about that
[19:12:58] <ebernhardson>	 they should, -internal is for automated queries from wikibase afaik
[19:14:10] <inflatador>	 ACK
[19:20:35] * ebernhardson ponders how the backfill release hack is quite odd in k8s land...may revisit that some day after we have rate limiting guarantees
[19:20:57] <ebernhardson>	 would fix some uncertainty around deploy/destroy of the backfill releases
[19:23:05] <inflatador>	 You say "hack", I say "clever workaround" ;P
[19:26:29] <ebernhardson>	 it does what we needed at the time, but yea :)