[07:02:37] surely to look at, it was in our long-term backlog to automate that stuff ;) [07:52:46] <_joe_> I was always doubtful of the cost/benefit ratio [08:23:17] I'll ask you back your feedback after this week of clinic duty :-P [08:23:33] but hey, if it's already done... might be worthed anyway ;) [08:37:55] <_joe_> that's my point [08:38:07] <_joe_> also it's not like I didn't have to do clinic duty in the past :P [12:30:09] godog: around the time of https://wikitech.wikimedia.org/wiki/Incident_documentation/20190723-network_eqiad#Actionables we had some ideas about associating rack/row with prometheus labels, did we file a task or do anything? I forget [12:33:31] cdanis: afaik no we didn't [12:33:40] (file the task that is) [12:34:26] did we get ratholed on how best to get the data from netbox into puppet? [12:34:30] is that what happened? [12:35:29] most likely :) [12:35:36] IIRC that's right, that's one of the missing steps for sure, "something" that given an hostname returns rack/row [12:35:50] ok [12:35:56] that sounds right, thanks <3 [12:36:11] where we need that info? prometheus hosts? [12:36:54] there are a few ways I think you could do it [12:37:34] with, I think, some easier ways available if we decide to only apply it to certain metrics (e.g. just for node_exporter network interface data) [12:38:05] yeah I remember we chatted about it at the team meeting, but nothing in phab yet afaik [12:38:22] ok! [12:38:52] at a meta-level I am wondering if we want a script that would turn a bunch of "blah blah (TODO: Create task)" bullet points into placeholder tasks, and then go back and edit the wiki page for you [12:39:40] because I've only read 5 pomos/indos this morning and I've already seen about 15 of those still (TODO: Create task) for months [12:41:36] yeah that and/or a template to link to phab [13:33:32] i think _joe_ noticed last week that yaml treats y/n, yes/no and on/off as boolean values. https://arp242.net/yaml-config.html is an intresting write up of some other pitfalls