[20:16:57] bearND: better not deploy now [20:18:00] Pchelolo: oh, want me to rollback? Only got the canary right now. [20:18:13] just abort it, not good timing [20:18:19] there's like 2 outages going on right now [20:19:15] Pchelolo: if the canary is successful do I just say "no" when it asks to continue? [20:19:43] oh well, I'll try that. [22:06:13] bearND: our outage is finally over, sorry. I don't think your deploy would've affected something, but who knows [22:07:15] also, there was an unrelated outage for appservers, so for the EU night we've decreased concurrency of changeprop significantly, just to be sure, so someone might notice that updates to mobile content are beeing propagated a bit slower [22:09:33] Pchelolo: well, we had a ton of changes. So, anything's possible. We have https://gerrit.wikimedia.org/r/c/mediawiki/services/mobileapps/+/444322 queued up and were going to test to see if it improves performance for the media endpoint. [22:09:49] Do you think I could try to deploy now? [22:11:08] bearND: if you're confident it will no cause another outage :) [22:15:28] Pchelolo: I think so, to a certain degree. [22:16:22] ok :) it seem to touch mostly media, so it should be fine I guess, everything is in fairly solid state now [22:27:14] Pchelolo: deployment is done. So far, so good. [22:27:21] great :) [22:29:12] But I guess there is not much usage in the media endpoint yet anyways.