[12:38:14] lmata: ^ is that "devnull" meant to be there or is is a bot's bug? First time I have seen that [12:40:42] marostegui: tthere is an override for Andrea AFAICT today that is taken by devnull [12:42:53] volans: yeah, but I never saw devnull on the topic, hence my question [12:43:02] maybe I just missed it when it happened [12:43:12] eh [12:47:08] marostegui: this is right. i think d.enisse|m is OOO today (MX local holiday) and has an override for /dev/null and is on-call today./.. so the /dev/null override contact is showing instead [12:47:37] so basically what volans said [12:47:55] :D [12:49:06] would it be helpful to add to the label of /dev/null contact to be more explicit? [12:49:29] lmata: nah, it is ok, I just never saw that before, so I was wondering if that was expected or not [12:49:31] thanks [12:50:07] lmata: so what's the expected outcome if one of the two members of the oncall isn't available (ie public holiday), just have one person oncall for that shift? [12:51:46] I guess we should omit devnull in the topic, it's just confusing [12:51:50] and surely not someone oncall :D [12:52:44] "Your feedback is very important to us. [12:52:54] " --- puts feedback form into shredder [12:56:34] marostegui: indeed, we have 2 folks in business hours, but if its a hoilday (in your country), its not really business hours. [12:56:34] Ideally we would try to find folks to cover that day within reason (see https://office.wikimedia.org/wiki/SRE/OnCall-FAQ), in this case I wasn't aware of the MX holiday until yesterday, but I'm attempting to find someone this morning anyways [12:58:20] lmata: got it, thanks! [13:06:14] ツ [13:10:32] actually a pretty good forum to advertise... If anyone in the Americas can cover an on-call shift today please reach out to me. Thanks in advance! [14:00:04] lmata: I can cover today if still needed [14:00:26] <3 thanks! it is needed. much appreciated [14:45:51] mutante, apergos, can you figure out what's happening with this dump? It looks like we're expecting a phab100x host to be running rsync but I can't tell if that was ever true. https://phabricator.wikimedia.org/T322221 [14:55:23] rsync server, I mean [16:21:47] rzl: is this you??!? https://what-if.xkcd.com/161/ [16:30:14] Is there a way to get the prod puppet ca cert publicly? (so we can validate certs signed with it) [16:30:27] andrewbogott: I replied on the task, phab1004 (?) probably should be doing that now and it's been true in the past that phab1001 did so [16:30:30] -rw-r--r-- 1 root root 1733902681 Oct 14 03:28 phabricator_public.dump [16:30:35] last dump we got [16:30:50] I saw, thanks. I bet the rsync server was never puppetized [16:31:02] that I really don't know... [16:33:52] ottomata: yep! [16:34:40] rzl: you rock, really nice reading [16:35:55] the whole book is great if you haven't checked it out yet :D [16:36:19] dcaro: 301 to jbond / moritzm ;) [16:37:09] uh. just to be 100% clear. the question was me, the answer was written by the author, Randall Munroe [16:37:52] a.k.a xkcd :D [16:41:26] rzl: yes yes but I didn't really think about that question and it is interesting :D [16:41:52] whew :D [16:41:59] dcaro: there is https://apt.wikimedia.org/wikimedia/pool/main/w/wmf-certificates/ which has both the pki and puppet ca [16:42:24] there is also https://github.com/wikimedia/puppet/blob/production/modules/profile/files/puppet/ca.production.pem which is just the puppet ca, we could probably put it $somewhere elses as well if needed [16:42:55] also the wmf-certificates package is deployed to the whole production (thanks to John :) [16:43:36] thanks! quite useful