[09:55:52] should membership in the wmf ldap group still grant RO access to things like netbox (like wikitech says) or has something changed since moving to use CAS? [09:57:39] yes it should [09:57:51] in general, do you have a specific use case? [09:59:07] because IIRC there is still an issue in netbox with users that gets created with a hash username, but you should ask slyngs ( https://netbox.wikimedia.org/users/users/?sort=first_name ) [09:59:17] that's for me, when I try to login to Logstash or Grafana I'm getting a `Authentication Failure Service access denied due to missing privileges.` [09:59:38] even though I'm part of `wmf` ldap group [10:00:48] ops, netbox-readonly-access and wmf should grant Netbox access [10:00:49] same for netbox and libreNMS [10:01:27] Logstash requires nda, logstash-access or ops, but not WMF [10:04:21] matthieulec: try logging out and back in at idp.wikimedia.org? [10:04:27] matthieulec: You might need to sign out of CAS-SSO / idp.wikimedia.org and sign back in, if you haven't done so after getting wmf access [10:06:10] ok yes it solved the issue [10:06:12] thanks! [10:06:29] Sorry, we really should find a way to make that clear :-) [10:06:46] slyngs: automatically logging out peoole when groups are changed? [10:07:00] That's a good idea [10:07:16] Simple too [11:19:13] test-cookbook is failing in cumin1002 with sudo: cookbook: command not found [11:19:31] what am I missing? [11:19:32] vgutierrez: cumin1002 is about to be decommissioned I believe [11:20:28] indeed (see tread from moritz [sre] New Cumin host: cumin1003.eqiad.wmnet ) [11:20:59] what's the patch? not sure how the test-cookbook is still using it [11:21:06] test-cookbook CI [11:21:41] cumin1002 no longer uses the cluster::management role, but maybe it's still hardcoded in some CI image [11:24:04] moritzm: I think that valentin was trying to use test-cookbook manually [11:25:26] ah, ok [11:25:43] then the fix it simply to use cumin1003 ofc :-) [11:26:47] yep [14:56:46] dpogorzelski: ok to merge your puppet change? [15:04:37] dpogorzelski: another ping, your puppet change https://gerrit.wikimedia.org/r/c/operations/puppet/+/1206402 was never puppet-merged [15:05:18] maybe klausman is around? [15:05:40] let me look [15:06:04] yes, go ahead and merge it [15:06:15] thanks, doing [15:32:47] FYI, I'm going to start getting set up for today's work on T352245, which will involve two etcd restarts in codfw (which will happen closer to 16:00). [15:32:47] the work I'll be doing prior to that involves making those restarts less impactful (e.g., temporarily moving etcd replication to another node). [15:32:47] cc: Raine jhathaway [15:32:47] T352245: Migrate the etcd main cluster to cfssl-based PKI - https://phabricator.wikimedia.org/T352245 [15:33:03] I'll be noisy in -operations as things progress [15:33:08] thanks swfrench-wmf [16:29:53] mutante: good to merge your patch? [16:30:59] mutante: this would be https://gerrit.wikimedia.org/r/c/operations/puppet/+/1204982 [16:33:44] never mind - sorted in -operations [16:48:16] FYI, I'll be poking around a bit to double-check some things, but my work is ~ done [18:24:10] my ganeti host reimage is failing due to a puppetdb looking; is this familiar to anyone? [18:24:13] https://www.irccloud.com/pastebin/eLR9Y22x/ [18:25:13] this is a second attempt after a failed install so I assume it's in some unanticipated halfway state [18:26:44] andrewbogott: does it have that "profile::puppet::agent::force_puppet7" in Hiera ? [18:27:17] it does, but also that check should just warn me and wait for me to add it if it's not there [18:27:24] https://gerrit.wikimedia.org/r/c/operations/puppet/+/1206932 [18:30:32] andrewbogott: 2 things.. there was just an actual networking issue in that time.. so if the issue was it could not reach the puppetdb it might be incredible timing [18:31:37] and the other.. ..never mind.. best I have is .. repeat it and see if it happens twice first [18:32:21] it happens more than twice [18:32:43] of course if I set --new to force creation of the puppetdb entries it says they're already there and removes the --new [18:33:21] andrewbogott: which VM? [18:33:37] it's brand new, has never worked properly [18:33:40] sudo cookbook --no-locks sre.hosts.reimage --os trixie cloudidp2001-dev --new --force [18:33:52] fqdn is cloudidp2001-dev.wikimedia.org [18:34:22] your site.pp entry is broken: https://phabricator.wikimedia.org/P85370 [18:35:26] bah, that's what I get for copy/paste [18:35:27] thx