[07:41:52] XioNoX, topranks o/ - I was checking the network-related icinga alerts and the current ones (cr3-ulsfo <-> cr2-eqord <-> cr2-codfw) seem to be related to Telia's maintenance. But I noticed that we have also other two maintenance windows overlapping, Lumen and NTT. I didn't check in depth what links are affected, but if you could confirm that nothing has the potential to explode I'd be grateful :D [07:42:56] (I guess that NTT is transit and not transport so I nothing really problematic, in theory) [07:44:23] and Lumen is again eqiad <-> esams IIUC, so nothing new [07:44:33] (https://phabricator.wikimedia.org/T287469) [07:44:46] elukey: thanks for checking! [07:45:37] sorry for the ping but I got worried, it seems all good [07:45:43] at least from my understanding :D [07:46:05] elukey: thanks! I think I noticed this was coming up and came to same conclusion you've put above, but let me dig into it :) [07:46:21] And never a problem to ping us, always better to be safe than sorry! [07:46:47] <3 [07:54:05] majavah: nice work on mediawiki-extension-codemirror! https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=994145 [07:54:14] majavah++ [07:55:53] majavah: indeed! [08:08:55] <3 [09:01:40] <_joe_> majavah: you know you're now part of our plan to take over debian, right? [09:02:07] <_joe_> moritzm: re: php upgrades, when you upload it to apt.wikimedia.org I'd rebuild all php 7.2 docker images [09:02:33] <_joe_> and we need to remember of this patch once we move to 7.3 [09:02:40] <_joe_> (I hope not so far in the future) [09:07:23] ack, I'll wait for Subbu's test of the backport, then I'll ping you when to rebuild the images [10:28:53] _joe_: I'm already doing my part with that package? :D [13:53:12] In around 1h we'll have wikitech in read only move for a few minutes for a migration (https://phabricator.wikimedia.org/T167973) [14:46:29] In around 15 minutes we'll have wikitech in read only mode for a few minutes for a migration (https://phabricator.wikimedia.org/T167973) [14:57:30] * bd808 is excited for wikitech db move :) [14:57:38] \o/ [14:59:49] * volans starts editing wikitech :-P [14:59:57] marostegui: anything we can help with? [15:00:07] volans: not for now, thanks! [15:02:35] let's coordinate on -operations [18:10:49] _joe_: I happen to be doing some unrelated changes in Logger and am seeing that before logstash and before udp2log, the debug/warn/error output from MW did actually use wfErrorLog/error_log in some distant past at WMF. [18:11:27] its' only the current iteration with logstash/kafka that involved syslog at all [18:40:31] Wanted to check, does anything in particular happen when adding a new cluster to wikimedia_clusters in the common hiera? About to add a new cluster and just curious about side effects [18:47:48] ebernhardson: would definitely like to hear from other SREs, but looks like this is a pretty generic thing that the prometheus part of puppet uses to identify servers running apache and memcached, pybal uses to get a list of mediawiki appservers, etc [18:48:38] I suspect it's also used by `conftool` to select by cluster like the command here: https://wikitech.wikimedia.org/wiki/Conftool#The_tools but not totally sure