[14:04:00] heh [14:04:13] reprepro somehow has grafana 6.3.4 for both jessie and stretch [14:04:28] I did not do this [14:04:44] I'm guessing at some point there was a reprepro update without a --restrict, or something? [14:07:15] that is logged on email, so would be easy to search [14:07:38] yeah just found it [14:07:40] Date: Mon, 02 Sep 2019 08:16:28 +0000 [14:07:42] reprepro changes: [14:07:44] replace stretch-wikimedia deb thirdparty/grafana amd64 grafana 6.3.4 5.4.2 -- pool/thirdparty/grafana/g/grafana/grafana_6.3.4_amd64.deb -- pool/thirdparty/grafana/g/grafana/grafana_5.4.2_amd64.deb [14:14:18] anyway, how we do packaging/upgrades here has never made much sense to me ;) [14:26:52] agreed, we've been chipping away at parts of the larger "debian packaging/repositories pipeline" problem but still ways to go heh [14:29:59] "be careful what you wish for..." or you will end up in charge of that project :-P [15:09:45] I have a silly question: I have a task X that is configured on puppet, and to check it has been done I also tell puppet to check X is completed succesfully or alert. [15:10:12] my question is how to check task (and its check) are not removed accidentally from puppet? [15:10:35] so we silently stop checking X completed [15:10:46] apergos: have a minute to give https://gerrit.wikimedia.org/r/#/c/operations/mediawiki-config/+/543664/ a look? (That's the much-discussed 'what is a dblist for?' patch from yesterday) [15:15:37] lemme see [15:16:59] ugh I forgot to ask marosteg ui about this (but I pinged about something else and he's very busy for the next couple days) [15:23:08] the patches look ok as far as they go, but I have a more basic question [15:23:30] which probably just exposes my ignorance but here goes anyways [15:23:58] why is this wiki in production config files at all, if it's a lab test wiki and will be running now on a labs instance? [15:24:15] (I have a meeting in 6 minutes, be forewarned) [15:24:37] s/labs/cloud/ but you get the point [18:28:50] apergos: it runs on bare metal, not on a VM [18:28:55] behind the normal caches, etc. [18:37:54] ah I was completely thrown off by the now very sensible naming scheme! [18:38:09] yes I see it is a regular prod host... then I think, I *think*, that everything is good to go [18:38:28] but gee I would love to have someone else swear to it first that is more sure of the db config setup [18:38:31] I'll +1 it though [18:39:14] and done [18:39:39] apergos: any thoughts about who that someone else might be? [18:40:36] well tbh either maroste gui or maybe um ... anomie? (not sure) [18:41:36] it's just, as much as I've used this data to read from and put queries together, i've not actually changed it so... [18:42:31] yep, me neither, not recently [18:45:27] anomie is online now in #wikimedia-dev [18:45:40] though I don't know if they are actively paying attention to irc [18:46:03] at the moment I can't get apache2/php to work at all on my server so can't test my patch locally yet :( [18:46:13] ouch [18:58:00] andrewbogott: LGTM! note that it is totally fine to take over an mwdebug host for testing, just by scp'ing these files to it. just !log you are doing so beforehand (and when you run scap pull to reset it) [22:47:53] https://hackaday.com/2019/10/16/pack-your-bags-systemd-is-taking-you-to-a-new-home/