[06:51:58] <_joe_> bd808: I think the collab team owns mailman, but I might be wrong [07:06:41] _joe_: yes we do! adding our team to the task [15:09:06] bd808: I spent a little time looking at that issue yesterday, I'll see if I can crack it this morning [15:58:28] Thanks for the mailman/spamassissin responses folks. :) [16:34:01] urandom: I see aqs1012 and aqs1023 marked as Active in Netbox but reported to have disappeared from PuppetDB and I don't see any open task for them, is that known? [16:42:30] volans: yes, both are connected to a refresh. 1012 failed and was prematurely replaced when it failed, and 1023 was insetup and being reimaged when it tripped over a switch bug [16:42:33] https://phabricator.wikimedia.org/T412830 [16:43:07] do they have an OS right now? [16:43:08] so 1023's reimage will be resumed, and aqs1012 will go through the whole decommission rigor when the others are [16:43:15] are they powered up? [16:43:39] 1012 is unbootable [16:43:46] well, both are unbootable [16:43:52] they are ghost hosts, so should not be up with an OS if not in puppet, and if they are broken should be marked as such in Netbox [16:46:17] neither is up with an OS [16:47:36] then let's mark them as failed in netbox I'd say [16:47:50] they are reported by a netbox report, hence the ping ;) [16:48:32] so aqs1012 is coming down (eventually), so should that be decommissioning? [16:48:49] that will be taken care when you'll run the decom cookbook [16:48:50] and will doing so impact the cookbook when the time comes? [16:48:55] so failed. [16:49:00] yes or run the decom :D [16:49:11] depending what are your hopes :D [16:50:23] I didn't do the decomm of aqs1012 because I generally do them all at once for a refresh [16:52:06] which should have been done a long time ago.. [16:54:00] (for posterity, both are marked as failed now) [16:54:43] thanks [17:42:57] I realize it's friday, i'd like to ship a mw-config patch that changes a test search cluster from http to https. Checked with releng already, no concerns. This is just a misconfiguration due to variation in how this new cluster works vs existing: https://gerrit.wikimedia.org/r/c/operations/mediawiki-config/+/1248858 [17:43:18] it's not used for any prod request flow, special debug parameters are required to invoke this cluster [17:44:59] ebernhardson: as I'm oncall right now, I think it's ok given the very specific nature of the change and limited impact [17:45:05] thanks! [17:45:40] ping if you need anything from us