[09:41:45] Hi! Is anyone around? I'm having some issues SSHing to a new horizon instance. I can connect to bastion and other projects no problem, but this is the first I've set up from scratch so I've probably missed something. Help:Access says to look for "“Finished puppet run”, BEGIN SSH HOST KEY FINGERPRINTS..." in the log, and I'm not seeing that. The project is wikicitevis :) [10:10:18] hi Samwalton9 [10:11:05] which is the error you see in your ssh client? [10:11:44] i.e, connection refused, timeout, etc [10:15:07] Permission denied (publickey) [10:15:46] I get it when going via the ProxyCommand method, or when connecting with agent forwarding via bastion (I can connect to bastion no problem) [10:16:18] 2018-07-05T10:13:14.779200+00:00 wikicitevis-prod nslcd[432]: [45e146] (re)loading /etc/nsswitch.conf <--- I see this [10:19:18] I'll be honest I'm not sure what that means :D [10:23:40] same for me :-) [10:25:19] All I've done is set up the instance, add a web proxy, and tried to ssh via the two methods outlined in Help:Access [10:25:30] ok [10:25:34] I'm checking [10:25:38] Thanks :) [10:27:46] I can't jump using ssh either :-/ [10:29:07] [ 22.463321] rc.local[385]: + puppet agent --onetime --verbose --no-daemonize --no-splay --show_diff --waitforcert=10 --certname=wikicitevis-prod.WikiCiteVis.eqiad.wmflabs --server=puppet2018-07-05T09:04:00.398176+00:00 wikicitevis-prod rc.local[385]: + puppet agent --onetime --verbose --no-daemonize --no-splay --show_diff --waitforcert=10 --certname=wikicitevis-prod.WikiCiteVis.eqiad.wmflabs --server=puppet [10:29:07] [ 23.097142] rc.local[385]: [1;31mError: Could not initialize global default settings: Certificate names must be lower case; see #1168[0m [10:29:15] ok, this could be a clue [10:29:34] puppet didn't run because the project name [10:29:40] which must be all lowercase [10:30:00] we may require to delete the project and create it again [10:30:42] ahh [10:30:44] That [10:30:48] *That's no problem [10:32:55] Samwalton9: deleting instances and project now [10:33:03] Thanks! [10:35:34] did you create a proxy? I think I forgot to delete it [10:35:44] Yeah [10:36:00] and I deleted the project already... hopefully openstack is smart enough to delete related resources [10:36:04] by himself [10:36:08] or no? no idea :-) [10:36:28] by itself* [10:36:34] * Samwalton9 shrugs and hopes [10:39:06] Samwalton9: try again, I just created the new project [10:43:24] Starting up a new instance, seeing a lot of log spam about certificate issues, not sure if that's expected or not [10:46:32] * arturo looking [10:47:39] doesn't look good [10:47:44] let's wait a bit more [10:48:16] not sure how reliable is removing a project and then creating a new one with the same name just modifying the case [10:49:07] in the meantime, we could create a `wikicite-vis` project :-P [10:49:25] 'bis' [10:49:50] Hauskatze: is currently named `wikicitevis` [10:50:06] then wikicitevis-bis :D [10:50:10] yeah that's fine, the project name doesnt matter that much :D [10:50:53] Samwalton9: let's wait for 5 more minutes? or we could wait for andrewbogott to be awake in case he has some other idea on how to fix this [10:52:59] I see labslogbot created on wikitech a page called 'wikicitevis' so as far as I can see the instance is correctly created [10:53:26] wikicitevis-prod.wikicitevis.eqiad.wmflabs [10:55:07] arturo, if you think renaming the project would work then I'm fine with that, but if you'd like to fix the problem more fully I'm happy to wait, there's no rush! [10:57:52] Samwalton9: on openstack I see the instance is called wikicitevis-prod, can you try to connect there maybe? just guessing [10:58:11] https://tools.wmflabs.org/openstack-browser/server/wikicitevis-prod.wikicitevis.eqiad.wmflabs [10:58:40] Yeah I'm trying to connect to wikicitevis-prod.wikicitevis.eqiad.wmflabs and still getting the same issue [11:08:19] there is probably some issue in the labs-puppetmaster server with the renaming [11:09:36] but I see the cert in the puppetmaster [11:09:36] "wikicitevis-prod.wikicitevis.eqiad.wmflabs" (SHA256) D7:A0:FB:DD:52:B6:82:67:8F:CD:F2:B1:88:68:9D:36:12:E8:B4:66:94:5A:79:CE:35:48:B7:97:89:FC:80:A8 [11:09:37] D7: Testing: DO not merge - https://phabricator.wikimedia.org/D7 [11:11:44] I just ran `aborrero@labpuppetmaster1001:~$ sudo puppet cert sign wikicitevis-prod.wikicitevis.eqiad.wmflabs` [11:11:50] and things are now moving [11:12:26] ooh [11:12:27] Samwalton9: can you connect now? [11:12:49] Same error :( [11:12:55] ok wait [11:13:01] I will clean the certificate [11:13:08] please, remove the VM and create it again [11:13:44] Samwalton9: ^^^ [11:14:16] Launching... [11:15:02] hey andre__ :-) [11:16:28] Samwalton9: puppet seems to be doing actual things now [11:16:34] Hooray! [11:17:16] Samwalton9: try now! hopefully it will work this time :-) [11:17:21] I'm connected :D [11:17:26] :-D [11:17:26] Thank you so much! [11:17:27] great! [11:17:32] thanks you Samwalton9 !! [11:17:58] I'll make a note of this on the original project request task [11:18:27] ack [11:19:00] I also added a note in our docs: https://wikitech.wikimedia.org/w/index.php?title=Portal:Cloud_VPS/Admin/Projects_lifecycle&diff=1796215&oldid=1796056 [11:19:19] Awesome :) [11:19:32] * arturo feeling useful today *^^* [11:23:17] * andre__ waves [12:15:00] !log wmam deleting the wikikids node due to request from WMAM and WMF legal [12:15:01] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Wmam/SAL [13:35:25] arturo, Samwalton9, did you delete all the VMs in the project before deleting the project? [13:35:51] andrewbogott: I would say so [13:36:01] Im not sure about the proxy [13:36:35] and the project moved from lowercase to MixedCase? or the other way around? [13:36:41] it is possible to have orphaned VM? [13:36:50] from mixed to loser [13:36:53] lower* [13:37:05] arturo: yeah, I think it is possible to orphan a VM [13:37:16] but I'd expect all lowercase to work better than CamelCase [13:37:29] Are there still problems with the puppetmaster or is it mostly better now? [13:37:38] it should be solves now [13:37:45] solved* [13:38:01] ok then :) [13:38:07] the issue with mixed case was with the puppetmaster [13:38:15] (certificate issues) [13:38:27] not with openstack itself [13:41:02] ok, that's (potentially) less complicated [15:16:25] hello friends! I am trying to delete a continuous grid job with `qdel`, but it's a very resilient job that's just refusing to die... can an admin force kill it for me? the ID is 243564 [15:16:50] musikanimal: Like... murder it? [15:17:02] Also, hi! Long time no chat, musikanimal :-) [15:17:40] maybe humane euthanasia, but if it comes down to a brutal murder, than so be it [15:24:35] musikanimal: I force deleted 243564 for you [15:24:49] thank you!