| mwhudson | whuh | 01:14 |
|---|---|---|
| mwhudson | maas-cluster-celery is failing to start for me | 01:14 |
| mwhudson | sl = self._semlock = _multiprocessing.SemLock(kind, value, maxvalue) | 01:14 |
| mwhudson | OSError: [Errno 13] Permission denied | 01:14 |
| mwhudson | [2013-04-24 02:13:58,398: INFO/MainProcess] process shutting down | 01:14 |
| bigjools | wha | 01:15 |
| mwhudson | dailybuilds ppa installed on precise | 01:15 |
| mwhudson | seems to be inside celery's start up code :( | 01:16 |
| mwhudson | python -c "import multiprocessing.queues; multiprocessing.queues.SimpleQueue()" fails the same way | 01:17 |
| mwhudson | hm, seems to be /run/shm permissions | 01:19 |
| * mwhudson tries rebooting the server (needs doing anyway) | 01:22 | |
| mwhudson | ok that fixed it | 01:26 |
| mwhudson | bigjools: argh | 02:43 |
| bigjools | ruh roh | 02:43 |
| mwhudson | bigjools: pserv doesn't appear to be serving a pxelinux.0 file | 02:43 |
| bigjools | did you import images? | 02:44 |
| mwhudson | yes | 02:44 |
| mwhudson | but i had that celery oddness | 02:44 |
| bigjools | arm is a bit different here | 02:44 |
| mwhudson | so maybe something didn't happen that shoulda orta | 02:44 |
| bigjools | mebbe | 02:44 |
| mwhudson | the "region controller doesn't know about images" went away eventually though | 02:45 |
| mwhudson | re run the import pxe script? | 02:45 |
| bigjools | mmmm if that's gone then it;s imported them | 02:47 |
| bigjools | but some of it could have failed | 02:47 |
| bigjools | did you run it manually or use the UI? | 02:47 |
| mwhudson | manually | 02:47 |
| mwhudson | oh | 02:47 |
| mwhudson | right, so the symptom i have of this is that the automatic pxe boot fails in uboot | 02:48 |
| mwhudson | if i set ipaddr and serverip and run pxe get; pxe boot it works and the node gets commissioned | 02:48 |
| bigjools | trying to remember how arm works ... | 02:48 |
| mwhudson | so it's really just pxelinux.0 not being there that trips it up | 02:48 |
| bigjools | it ignores that though, I thought? | 02:49 |
| mwhudson | mwhudson@lava-leg01:~$ tftp localhost | 02:49 |
| mwhudson | tftp> get pxelinux.0 | 02:49 |
| mwhudson | Received 26837 bytes in 0.0 seconds | 02:49 |
| mwhudson | what | 02:49 |
| mwhudson | adsadasdasdsada | 02:49 |
| bigjools | it just uses the location of it as a base directory for eveything else | 02:49 |
| mwhudson | right | 02:49 |
| mwhudson | but if it returns file not found it complains | 02:49 |
| bigjools | ok | 02:49 |
| bigjools | but you can get it as you just showed :) | 02:49 |
| mwhudson | i can /now/ | 02:50 |
| bigjools | however, the tftp server is sensitive to where the request comes from | 02:50 |
| mwhudson | probably because i re-ran maas-import-pxe-files | 02:50 |
| mwhudson | exceptions.AssertionError: No PXE template found in '/usr/lib/python2.7/dist-packages/provisioningserver/pxe'! | 02:51 |
| mwhudson | now i get that ^ | 02:51 |
| bigjools | vart da furk | 02:51 |
| mwhudson | oh | 02:52 |
| mwhudson | i think this is https://bugs.launchpad.net/maas/+bug/1064212 | 02:52 |
| ubot5 | Launchpad bug 1064212 in MAAS "If a machine is booted manually when in status "declared", TFTP server tracebacks" [Low,Triaged] | 02:52 |
| mwhudson | it's booting ok now | 02:52 |
| bigjools | could be yes | 02:52 |
| bigjools | maas has to know it's required to boot | 02:52 |
| bigjools | ideally we'd send something that shuts it down again | 02:53 |
| mwhudson | ok nodes are declaring themselves now | 02:55 |
| bigjools | yay | 03:06 |
| mwhudson | and now i can't get juju to work :) | 03:13 |
| mwhudson | (juju-go this time) | 03:14 |
| === rbasak-test is now known as rbasak | ||
| AskUbuntu | How to cat the `MAAS Key` of a user via `maas shell`? | http://askubuntu.com/q/284939 | 13:39 |
| === kentb-out is now known as kentb | ||
| rvba | Hi mgz, would you mind having a quick look at this when you'll get a chance: https://code.launchpad.net/~rvb/maas/empty-tag-bug-131232/+merge/160672 ? | 15:28 |
| rvba | I just want to know if it seems sane to you because you know that code probably better than I do. | 15:28 |
| mgz | rvba: sure | 15:42 |
| rvba | ta | 15:42 |
| allenap | roaksoax: Have you seen https://bugs.launchpad.net/maas/+bug/1172303? Do you know if this is just a difference between the packaging in quantal and raring? | 16:56 |
| ubot5 | Launchpad bug 1172303 in MAAS "MAAS_URL on maas_cluster.conf is quoted on quantal but not quoted on raring" [Undecided,New] | 16:56 |
| roaksoax | allenap: i think is packaging.. IIRC matsubara made a change recently and that might have changed the behavior | 16:57 |
| roaksoax | allenap: this might be the cause: http://bazaar.launchpad.net/~maas-maintainers/maas/packaging/revision/173 | 16:59 |
| roaksoax | matsubara: ^^ | 16:59 |
| roaksoax | allenap: yeah that's what has changed | 17:00 |
| roaksoax | matsubara: your fix broke something :) | 17:00 |
| roaksoax | allenap: we want it quoted right? | 17:05 |
| roaksoax | allenap: this would fix it: http://paste.ubuntu.com/5598853/ | 17:15 |
| allenap | roaksoax: What branch is used for the raring package then, if not that one? | 17:15 |
| allenap | roaksoax: That change in r173 looks like it's adding the quotes that the integration test is expecting to find, but it's odd that they're in quantal but not raring. I would have expected the other way round if anything. | 17:17 |
| roaksoax | allenap: that's the branch that's being used. I think the reason why its not being seen on quantal is because the affected change is in the SRU , with is in -proposed and not in -updates | 17:19 |
| roaksoax | allenap: so matsubara miught be comparing to that | 17:20 |
| roaksoax | allenap: oh hold on, what's on -updates doe snot have maas_url | 17:21 |
| roaksoax | allenap: so something has changed then | 17:21 |
| roaksoax | allenap: ont he dpkg-reconfigure it seems that the quotes never get added | 17:21 |
| allenap | matsubara: ^ I have to go now, but perhaps that'll lead to the root problem. Thanks roaksoax. | 17:22 |
| matsubara | thanks allenap and roaksoax | 17:23 |
| roaksoax | matsubara: i think this is the appropriate fix: http://paste.ubuntu.com/5598853/ | 17:23 |
| roaksoax | matsubara: manually applying this fix gives me: http://pastebin.ubuntu.com/5598866/ | 17:25 |
| matsubara | roaksoax, cool. so I just need to merge propose that against lp:~maas-maintainers/maas/packaging and packaging.quantal? | 17:26 |
| roaksoax | matsubara: and precise.sru | 17:26 |
| roaksoax | rvba: still around? | 17:28 |
| matsubara | roaksoax, ah ok. I'll mp it | 17:29 |
| allenap | roaksoax: I'm not sure that's the right fix. That changes the quotes from single to double, but the test is looking for single quotes, and the failure is that there are no quotes at all. | 17:29 |
| * allenap really goes. | 17:31 | |
| roaksoax | allenap: http://pastebin.ubuntu.com/5598887/ | 17:32 |
| roaksoax | allenap: yeah, so we should standarize and make the test not look for single quotes, but look for double quotes | 17:33 |
| roaksoax | because that's how everything is being done | 17:33 |
| matsubara | roaksoax, ok. I'll update the tests as well | 17:33 |
| roaksoax | allenap: the initial change was driven becuase a test was broken due to a leading whitespace | 17:33 |
| roaksoax | so we never noticed that there were no quotes in the first place | 17:33 |
| roaksoax | which would have been providing the above fix instead of the commited fix | 17:34 |
| matsubara | roaksoax, https://code.launchpad.net/~matsubara/maas/fix-MAAS_URL-quotes/+merge/160722 https://code.launchpad.net/~matsubara/maas/fix-MAAS_URL-quotes-quantal/+merge/160723 https://code.launchpad.net/~matsubara/maas/fix-MAAS_URL-quotes-precise/+merge/160725 | 17:42 |
| matsubara | I need to go for a dentist appointment and should be back in ~1h | 17:42 |
| === matsubara is now known as matsubara-afk | ||
| === cmagina_away is now known as cmagina | ||
| === matsubara-afk is now known as matsubara | ||
| === kentb is now known as kentb-out | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!