[01:55] <hobbyBobby> hope you guys like big files I'm uploading a 2Gb sos report
[09:59] <rvba> jtv1: I'm working on fixing the DHCP config writing code so that it copes with multiple managed interfaces.
[10:00] <jtv1> Great, thanks.
[10:25] <jtv> rvba: O
[10:25] <jtv> Ahem.  EKEYPOS
[10:25] <rvba> o_O
[10:26] <jtv> rvba: I'm updating ZoneGenerator to support multiple managed interfaces per nodegroup.  It may be relevant to what you're doing.
[10:26] <rvba> jtv: Is it?  I'm only working on the DHCP config.  Not touching the DNS side of things.
[10:27] <jtv> Ah of course.
[10:27] <jtv> No worries then.\
[11:32] <allenap> rvba: When I run the development Django service (make services/webapp/@run) it doesn’t /seem/ to be reading urls.py, but I don’t understand why. I can add `raise SystemExit(1)` at the module level and Django just rolls on. Any ideas?
[11:34] <jtv> ...Read on demand?
[11:34] <rvba> allenap: that's not what *I* see: http://paste.ubuntu.com/6825715/
[11:41] <rvba> allenap: btw, you remember our discussion about WithMACAddressesMixin.is_valid?  I was wrong, it's enough to populate form.errors to make the form invalid (and we actually have a test for this).
[11:44] <allenap> rvba: Cool, that’s great.
[11:46] <roaksoax> howdy all
[11:47] <jtv> hi roaksoax
[11:47] <roaksoax> allenap: so I consulted to various people and looked online, and it seems that having kmultiple setup.py's is a non standard thing to do, and that you usually should bundle verything under 1 single setup.py
[11:47] <roaksoax> jtv: howdy!
[11:47] <roaksoax> also *please* make sure whatever lands in trunk with new dependencies w2e are aware of
[11:48] <roaksoax> to have time to package it
[11:48] <roaksoax> MIR it
[11:48] <roaksoax> etc
[11:48] <roaksoax> we are pretty packed with work right now
[11:51] <allenap> roaksoax: They’re not all in the same directory.
[11:52] <allenap> roaksoax: The problem is that the setup.py at the top level does not align with a single Python package. It’s just a mess really.
[11:55] <roaksoax> allenap: i think you should consult with doko on this matter, or even barry
[11:55] <roaksoax> allenap: they will know better than me how you should package this in the python way
[11:57] <allenap> roaksoax: Okay, fair enough.
[11:58] <allenap> rvba: I’m going insane. I can do os._exit(1) at the top of urls.py and it doesn’t exit!
[11:58] <jtv> strace | grep urls.py?
[11:58] <jtv> Maybe you're just in the wrong directory or something.
[12:07] <rvba> allenap: looks like one of those cases where a stale .pyc is used instead…
[12:53] <jtv> rvba: is the DHCP change working out OK?
[12:57] <rvba> jtv: looks okay so far, but I'm not done yet and I'll also build a package and give it a spin in the lab to make sure it still works with only one managed interface.
[12:59] <jtv> rvba: an easy way to check that might be to diff the resulting configurations.
[13:21] <tomixxx> hi, i have set dhcp settings and reboot my maas-server and now i get an "internal server error" when i try to open 10.0.0.9/MAAS
[13:21] <tomixxx> any ideas
[13:22] <tomixxx> ?
[13:33] <tomixxx> noone? :(
[13:35] <rvba> tomixxx: please have a look at the logs in /var/log/maas and /var/log/apache2/*.log to see if you can find details about the error.
[13:36] <tomixxx> kk
[13:38] <tomixxx> in celery-region.log: i get Consumer: Connection Error : [Errno101] network is unreachable. trying again in 32 seconds...
[13:39] <rvba> Looks like a connectivity problem.
[13:39] <tomixxx> i have set the dhcp settings, oh and the network card interface is another one
[13:40] <tomixxx> and in /var/log/apache2 i get : error: Network is unreachable
[13:41] <tomixxx> do i need an active internet conneciton for maas?
[13:41] <tomixxx> because in my current setup, i have NO internet !
[13:41] <tomixxx> only a private network
[13:42] <rvba> Well, you need access to internet, either directly or through a proxy.
[13:42] <rvba> You need to download all the install images.
[13:42] <rvba> And the nodes need internet access to fetch the packages they need.
[13:43] <tomixxx> damn
[13:43] <tomixxx> i mean i have two network interface cards in my maas server
[13:43] <tomixxx> one is connected to the nodes in ne maas-network and the other to my uni-network
[13:47] <tomixxx> i dont know if one of the cards does not work...
[15:39] <BjornT> hi. i'm having an issue that sometimes when i juju deploy in a maas environment, cloud-init fails to execute the part-001 script. i couldn't see any other error message than error code 3, think. is this a known issue?
[15:44] <rvba> BjornT: I don't think it's a known issue, could you please file a bug describing the issue (https://bugs.launchpad.net/maas/+filebug)?  (make sure you include the result of running 'sudo maas dumpdata metadataserver.NodeCommissionResult' as an attachment to the bug.)
[15:48] <BjornT> rvba: sure. does it matter that i manually ran the script, which made the setup continue? i.e. will the NodeCommissionResult be different because of that?
[15:54] <rvba> BjornT: Do you mean that you ran the commissioning scripts manually?
[15:57] <BjornT> rvba: yes, i looked at which script cloud-init tried to run (/var/lib/cloud/instance/scripts/part-001) and ran it to see whether it would tell me how it failed. except that it succeeded that time.
[15:58] <rvba> BjornT: right, so the NodeCommissionResult table should contain the output the script when it failed.
[15:59] <rvba> the output *of* the script
[16:07] <BjornT> rvba: https://bugs.launchpad.net/maas/+bug/1273296
[16:15] <rvba> BjornT: Ta
[22:55] <bigjools_> BjornT, rvba: correct me if I'm wrong, but Bjorn is talking about cloud init, not commissioning scripts?
[22:56] <bigjools_> hey BjornT BTW :)