[06:16:59] thanks for your comments on the WCQS auth thread, gehel, ryankemper [06:55:15] mpham: 👍🏿 [06:55:52] it’ll have to wait till after new years but definitely tons of stuff for us to retro on as a team [07:03:51] agreed [07:04:29] maybe either as a normal team retro, or a project retro [09:53:38] gehel: by any chance are you around? [09:54:33] I've a question for you wrt elastic hosts and AAAA DNS records [09:59:58] it's about T271143#7582472 [09:59:58] T271143: Some Search Platform / Discovery clusters apparently do not support IPv6 - https://phabricator.wikimedia.org/T271143 [10:04:36] volans: looking [10:05:53] thanks [10:06:06] Wow, that was opened a long time ago! [10:06:17] mainly because if AAAA are not supported, when you put into production those new hosts they will breaj [10:06:20] *break [10:06:29] as clients will default to v6 connections [10:07:24] elastic should support IPv6, but let me check where it listens [10:07:42] accornding to Cas's comment: elastic - the nginx part seems to listen on ipv6, but the java process is not listening on ipv6. [10:07:47] but not sure if is still the case [10:07:54] and it all works fine [10:11:17] so... internal elasticsearch communication does not rely on DNS, but on internal cluster discovery mechanism. It is at the moment configured explicitly with the IPv4 of the host [10:11:40] external communication all goes through nginx, so should be good over IPv6 [10:12:12] that being said, we should configure IPv6 properly (or not... it does not really change much) [10:12:44] let me add a not to that ticket and tag it with [discovery-search]. Good first elasticsearch task for inflatador ! [10:12:52] :) [10:17:05] I vaguely remember Elasticsearc automatic IP discovery being broken in some way in our context. [10:18:08] volans: note that the es* servers are not ours [10:18:14] yeah I noticed :/ [10:18:37] btw for my PoV I'll keep the AAAA record assigned for some new elastic host I'm using to test some bios/idrac automation [10:18:44] so they are all the same with the new block [10:18:51] and then if needed we can clear the AAAA records all together [10:18:56] parent task updated [10:19:45] thx [10:20:31] * gehel needs to do some cooking [10:20:51] Oscar has invited 2 friends for lunch. I better be ready ! [10:22:19] :D [14:06:51] greetings! [14:10:50] morning [14:10:57] o/ [15:30:43] inflatador: not sure if you read the backlog. volans pinged us on an interesting task: T271143 [15:30:43] T271143: Some Search Platform / Discovery clusters apparently do not support IPv6 - https://phabricator.wikimedia.org/T271143 [15:32:58] Oh yeah, this looks cool. I assigned it to myself, let me know if I need to do anything else on the phab board [15:33:24] don't assign it to yourself until you're actually working on it [15:35:36] I can work on it now, unless you object (I'm just watching required onboarding videos). I can handle at least some of "If you could take some time to put any information you have about supporting IPv6 on these clusters" [15:36:03] inflatador: whatever make sense to you! [15:41:42] inflatador: you wiped the title [15:42:03] eh, upon reading the whole ticket, it looks like you and some others already did the discovery work. So, I unassigned the task. [15:42:09] RhinosF1 Ooops! Fixed [17:02:13] workout, back in ~30 [17:57:35] and back [19:16:24] lunch [20:07:36] and back [20:41:20] inflatador: fyi mailing lists have changed a bit [20:41:43] They no longer store passwords in plain text and have a million accounts. We have mailman 3 now [20:42:44] OK, I will update my onboarding docs accordingly. [23:30:06] I'm out, have a good one