[08:11:17] <atsukoito>	 good morning!
[08:12:52] <dcausse>	 hi!
[09:58:13] <atsukoito>	 hiya!
[09:58:43] <atsukoito>	 I'm now figuring out how to get to the beta cluster to run shell.php and debug T430882
[09:58:44] <stashbot>	 T430882: Beta cluster is not working due to error in Translate\TtmServer\ServiceCreationFailure - https://phabricator.wikimedia.org/T430882
[10:01:26] <dcausse>	 atsukoito: on the beta cluster it'll be "mwscript  shell.php --wiki testwiki" from deployment-mwmaint03.deployment-prep.eqiad1.wikimedia.cloud
[10:01:40] <dcausse>	 lemme check if you're in this project
[10:03:30] <dcausse>	 nop you don't seem to be allowed there, lemme try to add you
[10:04:53] <atsukoito>	 dcausse: thanks
[10:06:28] <dcausse>	 atsukoito: you should be in deployment-prep now (https://ldap.toolforge.org/user/atsuko)
[10:06:55] <atsukoito>	 yes, thanks!
[10:08:57] <atsukoito>	 yup, no ttm servers in the `$wgTranslateTranslationServices` indeed, will make a patch now
[10:12:48] <dcausse>	 hm... never saw a ttmserver index in deployment-prep, I assumed it was not enabled there
[10:13:26] <dcausse>	 but possibly some config shenanigans that forces us to have some things still defined there
[10:26:09] <dcausse>	 lunch
[10:28:20] <atsukoito>	 https://gerrit.wikimedia.org/r/1307099
[12:08:16] <dcausse>	 thanks!
[13:05:24] <inflatador>	 <o/
[13:18:05] <dcausse>	 o/
[13:23:37] <ebernhardson>	 \o
[13:26:23] <dcausse>	 o/
[13:38:00] <dcausse>	 cindy test runs in 2m on my machine but sadly only 4m on the cindy vm... I have an old intel but it probably runs at much higher freq than cloud vms... :/
[13:38:22] <ebernhardson>	 still pretty awesome! but yea servers tend to be slow but parallel
[13:39:00] <dcausse>	 on my machine it's flaky at 3 concurrent features, but can try what it does on cindy
[13:42:13] <ebernhardson>	 hmm, hard to say whats there.  cpuinfo reports haswell, but the last haswell cpu was released like a decade ago, seems like the hypervisor injecting some simple data
[13:43:41] <dcausse>	 yes, tried to look into this and see whether we could scale the freq but I doubt that's possible from a vm
[13:48:49] <dcausse>	 sigh... having an ophthalmic migraine that's building up, can't barely read characters on my screen, will take a quick pause, back in 30m hopefully
[13:53:47] <inflatador>	 I had one of those not too long ago, really freaked me out
[14:22:57] <inflatador>	 FYI, CODFW is done and I have repooled all 3 clusters there
[14:23:45] <ebernhardson>	 nice!
[14:24:53] <inflatador>	 Yeah, unfortunately we had the same quorum freak-out when it came time to kick out the last master on the main cluster, even after excluding it from voting.
[14:25:18] <dcausse>	 back
[14:25:41] <inflatador>	 Going down to 3 masters seems to work around this issue though...will add to docs
[14:25:51] <dcausse>	 nice, thanks!
[14:26:19] <dcausse>	 3 concurrent runs passed, going to push my luck a bit more :)
[15:23:49] <dcausse>	 5 seems ok, 6 started to fail, will perform a couple runs to see how stable it is
[16:30:42] <dcausse>	 dinner
[22:28:40] <quiddity>	 FYI, potential Search bug mentioned at https://en.wikipedia.org/wiki/Wikipedia:Village_pump_(technical)#Problems_with_search_with_multiple_keywords_with_wildcard_keys_(e.g._asterisk)_and_ending_with_A_and_S