[12:00:13] XioNoX, topranks: when you have 5 minutes I'd like to discuss what we want to do with the capirca script on netbox to avoid to remember to run it manually every few days. Do we need to validate the output? Should we run it automatically? Should we detect when there are changes that require an operator to run homer? [12:00:59] volans: the best would be to know on which host homer needs to be ran [12:01:07] but it's not possible in the current state of things [12:01:56] because the lists are used in different places not only the switches where the host is attached, right? [12:02:04] s/lists/definitions/ [12:02:54] for capirca, and for the network configs in general, yeah [12:03:33] there is no way to say "if setting X in changed in a yaml file, you need to run homer on device Y" [12:03:55] no, but I was wondering if the capirca script could know it [12:04:10] like making the diff from the last run and take the list of devices from the diff [12:04:20] but if say we use one of the definitions in the routers [12:04:24] than there is no way [12:05:53] yeah exactly [12:06:39] for now it would be nice to have a hook that runs the script when there is a change in devices [12:07:06] either device is provisioned/decom, or its IPs change [12:07:19] and same with standalone IPs, like VIPs [12:07:44] but doesn't solve the question of who runs homer [12:07:51] no [12:08:34] so the idea for now is "if you need an ACL change, run the capirca script and run homer" [12:09:52] should we keep the 3 days warning in homer's output? [12:10:03] and just ignore it in the daily email [12:11:12] we could remove it totally too [12:11:15] no strong preference [12:11:28] I don't think it's too much of an issue in the daily emails neither [12:11:47] we should discuss longer term network automation plans too [12:12:03] no need to do big changes until we know where we're going [12:12:08] up to you, no strong preference either, it might be useful to the people running homer to remind them that's outdated [12:12:41] but my subconscious would like to run the script when it sees the warning :D [12:13:34] hahhaha [12:13:54] excuse my lack of knowledge here - the warning is something homer produces if the capirca script has not been run recently? [12:14:06] topranks: exactly [12:14:11] WARNING:homer.capirca:Netbox capirca.GetHosts script is > 3 days old. [12:14:23] repeated multiple times btw [12:14:25] ok thanks [12:14:35] 7 in a random email I searched [12:14:39] volans: the fix is to not have any daily mails :) [12:14:52] ofc [12:14:54] :D [12:15:36] solid idea :) [12:16:13] I don't see it's presence in the mails to be a big issue anyway [12:18:20] * volans tempted to add a --quieter option to homer to log only errors :D [12:18:37] the timer already runs with --quiet that sets loglevel to logging.WARN [12:18:40] --kindofquiet [18:55:31] 10Puppet, 10Infrastructure-Foundations, 10SRE: Puppet does not undo manual "systemd mask $unit" - https://phabricator.wikimedia.org/T285425 (10Legoktm) [19:07:13] 10Puppet, 10Infrastructure-Foundations, 10SRE, 10Patch-For-Review: Puppet does not undo manual "systemctl mask $unit" - https://phabricator.wikimedia.org/T285425 (10Legoktm) [20:00:50] 10Puppet, 10Infrastructure-Foundations, 10GitLab (Initialization), 10Patch-For-Review, and 3 others: Puppetise gitlab-ansible playbook - https://phabricator.wikimedia.org/T283076 (10brennen) [21:39:12] 10Mail, 10Infrastructure-Foundations, 10SRE: Please create "grant@wikipedia.org" email handle to use for annual fundraising email test - https://phabricator.wikimedia.org/T285432 (10Peachey88)