| lathiat | Can anyone tell me if I can somehow login to a maas node running enlistment, to debug the enlistment not working? | 01:07 |
|---|---|---|
| lathiat | ah found it, backdoor | 01:12 |
| lathiat | Ah so my problem is maas is giving the DNS server for the region as it's primary interface, which goes through the router.. so the maas server is upset that the traffic is coming in and out different interfaces | 01:38 |
| lathiat | tada | 01:38 |
| mup | Bug #1471728 opened: MAAS web frontend hangs at Nodes listing if cluster not reachable <MAAS:New> <https://launchpad.net/bugs/1471728> | 09:51 |
| mup | Bug #1471728 changed: MAAS web frontend hangs at Nodes listing if cluster not reachable <MAAS:New> <https://launchpad.net/bugs/1471728> | 10:00 |
| mup | Bug #1471728 opened: MAAS web frontend hangs at Nodes listing if cluster not reachable <MAAS:New> <https://launchpad.net/bugs/1471728> | 10:09 |
| mup | Bug #1471731 opened: McDivitt console not available after deployment <MAAS:New> <https://launchpad.net/bugs/1471731> | 10:09 |
| mup | Bug #1471731 changed: McDivitt console not available after deployment <MAAS:New> <https://launchpad.net/bugs/1471731> | 10:12 |
| mup | Bug #1471731 opened: McDivitt console not available after deployment <MAAS:New> <https://launchpad.net/bugs/1471731> | 10:18 |
| ltrager | roaksoax_: morning | 15:58 |
| roaksoax_ | ltrager: morning | 16:01 |
| === roaksoax_ is now known as roaksoax | ||
| === med_` is now known as med_ | ||
| === liam_ is now known as Guest20028 | ||
| mup | Bug #1471946 opened: Installing on vivid yields a cluster not running because /var/lib/maas/secret is not there yet <MAAS:New for andreserl> <MAAS 1.8:New for andreserl> <https://launchpad.net/bugs/1471946> | 19:19 |
| === jfarschman1 is now known as MilesDenver | ||
| === jfarschman is now known as MilesDenver | ||
| bmorriso | I have a question about zones | 21:07 |
| bmorriso | It seems this is used for physical separation -- to ensure you don't provision a host in the same dc/rack/chassis/etc? | 21:07 |
| bmorriso | Is it possible to have a zone of a zone? So this chassis is in this rack, in this cage, in this DC? And I want MAAS to not deploy to a host that's in the same chassis, in the same rack, in the same DC... | 21:08 |
| roaksoax | bmorriso: no you can't have a zone of a zone unfortunately | 21:39 |
| roaksoax | bmorriso: but you can create tags | 21:39 |
| bmorriso | Yeah, I wondered about leveraging tags. | 21:39 |
| roaksoax | bmorriso: we would normally use tags for that kind of separation | 21:40 |
| bmorriso | Fair enough. I'm OK with that. We're still evaluating MAAS and figuring out how to use it and how it fits in our environment. | 21:40 |
| roaksoax | bmorriso: cool. Are you using 1.8 ? | 21:41 |
| bmorriso | Just upgraded last week. | 21:41 |
| bmorriso | Been using it since 1.6 | 21:41 |
| roaksoax | bmorriso: ah yeah, 1.6 is horrible :) | 21:41 |
| bmorriso | Got the job done, but yeah...My gripes about 1.8 are largely UI/UX | 21:41 |
| bmorriso | Also it seems pserv.log defaults to `/dev/null` o_O | 21:42 |
| bmorriso | But otherwise, it's treated me well. | 21:42 |
| roaksoax | bmorriso: not really. The log is now clusterd.log and regiond.log instead of pserv.log | 21:42 |
| bmorriso | Oh, I was looking at a config the other night, it said '/dev/null` | 21:42 |
| bmorriso | In 1.7 I had problems with pserv, so I was looking for that log in 1.8 | 21:43 |
| roaksoax | bmorriso: yeah we changed the config in 1.8. 1.9 will completely drop pserv.yaml, maas_cluster.yaml in favor of clusterd.conf and maas_local_settings.py in favor of regiond.conf | 21:43 |
| bmorriso | Awesome. Looking forward to it. | 21:44 |
| roaksoax | bmorriso: 1.9 will support both network and storage config as well | 21:44 |
| bmorriso | Any intentions of giving Curtin's docs more love in 1.8.x/1.9? | 21:44 |
| roaksoax | bmorriso: it is in the TODO. Definitely something we want to improve | 21:45 |
| bmorriso | I've managed to hack around most of my problems with help from this channel. But it puts my coworkers off who are less adventurous. | 21:45 |
| roaksoax | bmorriso: hehe, yeah curtin has really neat features but upstream (the maas team is not upstream of curtin) also lacks docs | 21:46 |
| roaksoax | bmorriso: so we are definitely looking to improve that for sure | 21:47 |
| roaksoax | bmorriso: and it is in the todo | 21:47 |
| bmorriso | No doubt. A one-page (sort of not-really-public) README leaves a lot to be desired. | 21:47 |
| roaksoax | indeed. But hopefully we can improve that this time | 21:48 |
| bmorriso | I actually have in the works a curtin installer that clones our ansible repo and runs a playbook to further the provisioning process. | 21:49 |
| wolverineav | 'Any intentions of giving Curtin's docs more love in 1.8.x/1.9?' +1 to that :) | 22:05 |
| wolverineav | on a related note, I'm currently copying some scripts over using wget in curtin. it works fine. the next step is to 'chmod +x /path/to/script.sh' - where it complains /path/to/script.sh doesn't exist. is there something that I'm missing? | 22:08 |
| wolverineav | I've tried to use the {{TARGET_MP}} and {{TARGET_MOUNT_POINT}} env variables, but its not available in the late_commands context | 22:09 |
| catbus1 | wolverineav: have you tried $TARGET_MOUNT_POINT? | 23:36 |
| wolverineav | catbus1: let me give it a try | 23:37 |
| wolverineav | ah, I seem to have hit a bug similar to the one seen before: https://bugs.launchpad.net/maas/+bug/1405998 my current maas.log: http://paste.ubuntu.com/11833449/ | 23:56 |
| wolverineav | # lsof -u maas | grep CLOSE_WAIT | wc -l - the output shows 984 | 23:56 |
| wolverineav | so i restart maas related services like dhcpd, regiond, clusterd, bind9 and then try again. and it works. | 23:57 |
| wolverineav | maas version is 1.8. should I create a bug? or it has already been handled in the trunk branch? | 23:58 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!