[10:08:43] I'm OOO today so can't do more on this but just a heads-up on https://gerrit.wikimedia.org/r/c/operations/deployment-charts/+/1287731 (cc kostajh) [10:21:27] hnowlan: is this urgent ? [10:25:26] I see, I am not sure I am confident to review this andn moste people are out [10:31:08] that doesn't look particularly worrisome to me if restricted to the liftwing endpoints, but we'd prefer to hold until Monday unless it's an emergency, I think [10:34:44] +1 [10:57:12] I’m not totally sure, but the failure rate seems low https://grafana.wikimedia.org/goto/afm4q4k1yejgga, so yes I would think dealing with this next week is fine [14:21:36] Dear serviceops, I would need your opinion on this ticket: https://phabricator.wikimedia.org/T424357 basically we have system of reducing TTLs on memcached in case of high lag, I want to remove it for reasons mentioned there. The only thing is that if we end up in a situation that two dcs are split and have a really high lag, would serviceops be okay to do rolling reboot of memcached after things are back to normal? we had such [14:21:36] incidents once or twice ever so it should be quite rare [14:24:57] dear amir, ok to discuss on monday? [15:02:21] effie: sure [15:08:16] Amir1: just tag us and potentially, add a comment of what is requested :)