[00:46:52] <legoktm>	 you all don't love writing man pages in groff format?
[06:49:50] <_joe_>	 legoktm: that will not be dignified with an answer
[06:56:32] <volans>	 s/f/s/g ;)
[08:03:49] <ema>	 volans: dignisied?
[08:06:08] <volans>	 ema: no, I was referring to 'groff' in lego's message ;)
[08:06:21] <ema>	 ah!
[08:06:33] <ema>	 I'm glad I asked, this would have kept me busy for a while otherwise
[08:06:38] <volans>	 ahahah
[08:06:55] <kormat>	 ema: this is why it's important to regularly express yourself. ;)
[08:06:58] <volans>	 yeah wasn't that clear I have to admit
[08:32:57] <_joe_>	 yeah took me a few secs too
[08:34:37] <XioNoX>	 or a sew secs?
[08:37:06] <ema>	 XioNoX: ha
[09:44:01] <apergos>	 how are we still stuck with nroff and troff decades later anyways?
[10:42:01] <jynus>	 this I think is an interesting read for sres, and 100% relevant to us: https://github.blog/2020-07-08-introducing-the-github-availability-report/
[10:45:37] <kormat>	 volans: how.. do i grant someone access to ssh into one of my cloud VPSes?
[10:45:51] <kormat>	 i can't even figure out how _i_ can log into it
[10:46:04] <jynus>	 kormat: familiar with horizon?
[10:46:18] <kormat>	 jynus: a little, unwillingly
[10:48:17] <jynus>	 actually, I don't think there is per-instance granularity, only per project
[10:50:32] <kormat>	 ahhh. AuthorizedKeysCommand /usr/sbin/ssh-key-ldap-lookup
[10:50:35] <kormat>	 never seen that before
[10:51:04] <jynus>	 but maybe you can play with sudo & groups, not sure what you want to achieve
[10:51:37] <kormat>	 jynus: i want to allow marostegui to ssh into a vps
[10:51:46] <jynus>	 which project?
[10:51:52] <kormat>	 automation-framework
[10:52:01] <kormat>	 is the only way to add him to the project?
[10:54:25] <jynus>	 for a global root I think people is quite generous with access
[10:54:32] <jynus>	 with reason
[10:54:45] <kormat>	 jynus: i don't see a way to do per-instance access v0v
[10:54:46] <volans>	 not sure if you are admin if not I can in a bit or any admin in the projecr in horizon can
[10:54:52] <kormat>	 volans: ok, done. what needs to happen now?
[10:55:14] <marostegui>	 I can access now
[10:55:16] <marostegui>	 Thanks guys
[10:55:23] <kormat>	 crazy black magic
[10:55:35] <jynus>	 so the granularity I meant was only if you wanted to give manuel only some rights
[10:55:43] <jynus>	 but I think won't be needed?
[11:17:39] <jbond42>	 hi all, have now migrated puppet production environment to using hiera version 5.  This should caus no observable difference however if you see anything strange involving hiera please  ping
[11:57:08] <_joe_>	 jbond42: \o/
[11:57:18] <_joe_>	 now we can seriously talk about environments imho
[11:59:45] <jbond42>	 yep, we can also support in module hiera defaults, although i dont know if thats something people want
[12:00:15] <XioNoX>	 is there a changelog I can understand? :)
[12:04:20] <kormat>	 XioNoX: _you_? no. ;)
[12:04:30] <jbond42>	 XioNoX: im not sure tbh, hiera 3 use to be a standalone gem but hiera5 is now part of puppet so i dont think a simple changelog exists.  At a high level there is not much functional difference.  however there are some minor difference for instance im pretty sure you can now define an undef default in hiera using `module::param: ~`.
[12:05:44] <XioNoX>	 well, that's good, one less thing for me to care about :)
[12:06:31] <jbond42>	 most of the difference are related to low level api design for instance compare git/puppet/modules/wmflib/lib/hiera/backend/nuyaml3_backend.rb with modules/wmflib/lib/puppet/functions/wmflib/expand_path.rb
[12:07:07] <jbond42>	 but yes you should be able to pretty muchy continue using things as is and as joe mentioned it opens up some new designes/functionality which we could start to make use of
[12:07:30] <jbond42>	 however wmcs would need to also migrate to hiera5 before we can serioulsy consider that
[12:07:58] <jbond42>	 probably/maybe ...
[12:34:56] <volans>	 jbond42: zero-prio, let's remember to update the onboarding chat slides ;)
[12:38:28] <jbond42>	 volans: ack will do
[12:39:31] <volans>	 thx :)
[16:17:48] <andrewbogott>	 XioNoX, jclark-ctr, cmjohnson1, I'm reviewing the switch/ceph/cloudvirt tickets and can't tell where we're at.  Are they just awaiting physical cabling?  Or are some of them already up and ready for OS install?
[16:19:14] <XioNoX>	 andrewbogott: from LLDP some servers are cabled, they need to be put in the proper vlan (defined in my comment) and then they should be able to be installed
[16:19:24] <cmjohnson1>	 They’re in need of a couple of things. I’m working on them
[16:19:45] <andrewbogott>	 is 'put in the proper vlan' a dcops thing or an arzhel thing?
[16:20:04] <andrewbogott>	 oh, nice — thanks cmjohnson1
[16:20:22] <XioNoX>	 dcops, but as they're new vlans, they might need my help
[16:21:01] <cmjohnson1>	 dcops thing.  Your cephosd servers will be ready today or first thing tomorrow barring any odd issues that pop up.
[16:21:09] <andrewbogott>	 woo!