[09:23:23] FYI I've created the pad for tomorrow's team meeting [09:56:21] thanks volans ! [11:35:20] anyone know how i can stop wmf_style lint complaining for https://gerrit.wikimedia.org/r/c/operations/puppet/+/531157 volans perhaps? [11:36:23] * volans feeling summoned [11:36:50] :) [11:38:37] mmmh it says [11:38:37] modules/role/manifests/spare/system.pp:1 wmf-style: role 'role::spare::system' should not include defines [11:38:51] and you're trying to add interface::add_ip6_mapped [11:38:52] there [11:40:25] yes this is a breach of the style but i want to breach it in this instance as i think it makes sense to have this here for spares and not in the multiple stanzas for spare in site.pp [11:40:35] i thought i could disable the check with # lint:ignore:wmf_styleguide [11:41:55] yeah me too :D [11:42:12] maybe the double space before # ? [11:42:15] didn't checkked the code [11:42:18] yet... [11:43:30] it normaly is 2 spaces i think my editor complaines with one, but also havn;t checked the code [11:45:13] i also tried https://phabricator.wikimedia.org/P8937 and https://gerrit.wikimedia.org/r/c/operations/puppet/+/531149/2/modules/role/manifests/spare/system.pp but they all fail [11:46:20] anyway if its not obvious dont worry about it i can look further just wanted to check i wasn't missing something [11:47:32] jbond42: I'm heading for lunch, I can have a look after it [11:47:35] if it's still an issue [11:47:47] the real author of that thing is j.oe btw ;) [11:47:49] * volans hides [11:47:51] thanks but dont worry about it its low priority [11:48:05] ack thanks [11:48:29] ttyl [11:48:35] later [11:57:04] <_joe_> jbond42: uh it's weird indeed [11:57:59] <_joe_> I smell a puppet-lint defect [12:00:56] hmm could be :(, Thanks for looking its good to know im not missing anything. Further im not sure this is a rabbit whol i want to jump down right now as this is not very important [12:01:28] ill maybe take afurther look on friday [12:21:32] jbond42: I think the issue has something to do with https://github.com/wikimedia/operations-puppet-lint-wmf_styleguide-check/blob/master/lib/puppet-lint/plugins/check_wmf_styleguide.rb#L374 [12:26:15] unfortunately the only way I've found to get it to pass is to wrap the whole class in a lint:ignore:wmf_styleguide block [12:28:08] thanks cdanis i really didn;t want to snipe anyone with this. i tried wrapping the whole class in https://gerrit.wikimedia.org/r/c/operations/puppet/+/531149/2. have just pushed another PS to move the ignoe:lint to line 1 see if that helps [12:28:23] yeah I tried ignoring line 1 and it did not help [12:28:39] I didn't go as far as looking into how the ignore functionality is implemented in puppet-lint itself [12:28:39] wrapping the whole file seems to work https://gerrit.wikimedia.org/r/c/operations/puppet/+/531149/3 [12:29:11] but I have suspicions that as because it's not mentioned in the puppet-lint "writing a plugin" docs at all, that it's buried in some inner abstraction [12:31:28] yes im gussing check_no_defines needs to be updated to specify the line with the offence instead of line: 1 but that is just a guess [12:32:29] I suspect it's harder than that -- a lot of the functions (e.g. check_hiera) iterate over tokens, which seems to be the expected interface [12:32:43] it's possible the ignore functionality omits ignored stuff from the tokens? [12:33:28] check_no_defines just looks at the final state, though [12:33:41] anyway, I'm done with this for now :) [12:37:18] :) thanks, I have another wmf-style lint thing in my mental backlog so ill take a better look when i tackle that [12:43:05] <_joe_> cdanis: uh in theory that helps not duplicating alerts [12:44:18] <_joe_> I suspect that the real issue here is puppet-lint itself [13:07:20] jbond42: I'm late to the party I noticed though that it was saying line:1 [13:07:25] * volans reading backlog [13:11:14] volans: indeed [13:11:19] https://github.com/wikimedia/operations-puppet-lint-wmf_styleguide-check/blob/master/lib/puppet-lint/plugins/check_wmf_styleguide.rb#L374 [13:11:44] ah [15:06:44] moritzm, jbond42: if we're adding mapped v6 IPs to the failoid hosts we should add the relative fw rules to v6 too I guess [15:07:09] although it's true that the fallback in gdnsd is v4 only at this time [15:07:21] I can assume that we'll add the v6 too at some point if it exists, dunno [15:07:24] thoughts? [15:07:55] volans: right now im just adding in the mapped v6 address. adding DNS will come later and i think thats when we will need to worry a bit more about firewall rules? [15:08:05] unless special case it's required only if the host has a AAAA record (when other servers might try to talk to it over v6) [15:08:29] jbond42: failoid is never used with names, just with hardcoded IPs as fallback in gdnsdc config [15:09:58] happy to add it to my CR but can you put a comment on the CR, i have raised a few of theses and few more to raise and may forget . [15:10:16] sure [15:10:23] XioNoX: sorry for all the reviews comming your way but its mostly just fyi will seek review from service owners [15:10:37] volans: thank [15:14:38] {done} [15:15:12] thx [15:19:19] jbond42: haha no pb at all :) [18:51:35] I tested Janitor locally and even if it's still quite rough, it's not that far from being usable in prod [21:22:40] haha https://phabricator.wikimedia.org/T230600#5418731