[03:32:15] Can we have a +2 on this quota request: T425905 [03:32:16] T425905: Quota increase request for project language - https://phabricator.wikimedia.org/T425905 [03:34:08] We also have 2 more quota requests pending [06:44:12] greetings [07:59:25] I'll kick off the work in T420565 [07:59:26] T420565: Audit tools memory requests vs actual usage - https://phabricator.wikimedia.org/T420565 [08:11:51] ok change merged, bumping version and deploying to toolsbeta [08:24:02] ok new webservice-cli is on toolsbeta now, I have verified that e.g. restarting fourohfour does the right thing [08:24:17] - memory: 256Mi [08:24:17] + memory: 128Mi [08:29:32] I'm test-running the roll-restart script on toolsbeta [08:35:31] ok deploying new webservice-cli to tools [08:45:59] will begin with restarting a sample of 10 tools https://phabricator.wikimedia.org/P92497 [09:05:56] I'm not seeing any disaster ensuing [09:07:42] I'll be resuming restarting webservices using default requests later today [14:04:50] dhinus, komla: horizontal scaling works for wiki-tts. So now they have a different, more possible quota request. https://phabricator.wikimedia.org/T425804 [14:17:19] andrewbogott: nice! it was worth asking in the end :) [14:18:00] yeah! [18:11:31] Not sure what is wrong. https://test.beta.toolforge.org/healthz is unreachable, but a webservice is currently running on the test tool. [18:58:32] Reedy: all traffic is failing isn't it? I'm going to try restarting the proxy [19:00:41] um... sorry Reedy, mistype, disregard [19:00:51] heh [19:00:53] The toolsbeta haproxies say "Server k8s-ingress-http/toolsbeta-test-k8s-gateway-2.toolsbeta.eqiad1.wikimedia.cloud" [20:56:58] LOL I did not paste the useful part of that log message! It is "Server k8s-ingress-http/toolsbeta-test-k8s-gateway-[12].toolsbeta.eqiad1.wikimedia.cloud is DOWN, reason: Layer4 timeout"