[10:05] <Daviey> smoser: RE: eth0.. not only is network adapters now consistently named.. you now can't guarantee they are called eth*.  biosdevname is now enabled.
[10:29] <rbasak> Daviey: would you object to: if bootif is supplied (on Intel via "ipappend 2"), then use that. Otherwise, if eth0 exists, then use that. Otherwise, fail.
[10:30] <rbasak> Incidentally, internally the netlink api does have an ordering. It is possible to ask the kernel for the name of the "first" interface. But what that is is undefined of course.
[10:34] <Daviey> well.. sounds reasonable.
[10:34] <Daviey> not quite worked out the ramifications tbh
[12:17] <roaksoax> jtv: howdy
[12:17] <roaksoax> jtv: so what's the status of MAAS DHCP?
[12:17] <roaksoax> rvba: ^^
[12:23] <rvba> roaksoax: hi.  jtv just put up for review two branches to add the writing of the interface config file.  We decided to start with that (i.e. do the same as for the DNS config) and get it working;  if it is possible to do the solution you suggested yesterday (fix the apparmor profile and have our own dhcp daemon) then what it will require on the upstream side will be very simple as all the config writing stuff
[12:23] <rvba> will already be in place.
[12:24] <roaksoax> rvba: ok, so those two branches are the ones that actually write the config to /etc/dhcp/dhcpd.conf?
[12:25] <rvba> roaksoax: no, that is already done.  The config file I'm talking about is /etc/default/isc-dhcp-server where the interface is specified.
[12:25] <rvba> The interface that the DHCP server should listen to that is.
[12:34] <roaksoax> rvba: ah I see, cause the version I released yesterday to quantal doesn't seem to be writing a dhcp config file
[12:37] <rvba> roaksoax: Jeroen will know probably better but I'll try and do some testing about that today.
[12:38] <roaksoax> rvba: ok cool
[14:06] <roaksoax> rvba: oh btw... was wondering if you guys were planning on working on the quantal support for maas for this cycle?
[14:06] <rvba> roaksoax: yep, it's scheduled.
[14:07] <roaksoax> rvba: is there an ETA?
[14:07] <rvba> roaksoax: Not that I know of.
[14:08] <roaksoax> rvba: ok thanks ;)
[14:08] <rvba> welcome :)
[14:14] <rvba> roaksoax: btw, this is bug 1013146 so if you have insights on how it should be fixed, don't hesitate to add comments on the bug.
[14:15] <roaksoax> rvba: will do, started looking at it yesterday but didn't really dig deep into it
[14:15] <rvba> roaksoax: ok, cool
[14:48] <roaksoax> rvba: so how do you think is best to enumerate the releases?
[14:49] <roaksoax> development, stable, lts, others?
[14:50] <rvba> roaksoax: should we simply name the releases using the result from 'distro-info --supported'
[14:51] <rvba> ?
[14:51] <roaksoax> rvba: right, but are we looking to support all of the releases?
[14:51] <roaksoax> rvba: or are we looking to support precise and up
[14:51] <rvba> roaksoax: I thought the plan was to support precise and up yeah
[14:52] <roaksoax> rvba: so how do you think we should be enumerating them here: http://paste.ubuntu.com/1187350/
[14:52] <roaksoax> how can we identify them
[14:52] <rvba> (I'm currently looking into the dhcp config writing pb btw)
[14:53] <rvba> roaksoax: how about storing the first release we support (precise) and then get the result of 'distro-info --supported' and parse that?
[14:53] <rvba> Do we really what to classify the releases like that (devel, lts) ?
[14:54] <roaksoax> rvba: if not, how can we make the attributes of the class be the release name itself?
[14:55] <roaksoax> rvba: it is not that we can do exactly as we do with the Architecture
[14:55] <roaksoax> that we can easily identify
[14:56] <roaksoax> rvba: unless for now we handle them like that, and then we find a better way
[14:57] <roaksoax> just to test quantal and precise
[14:57] <rvba> roaksoax: sounds good to me.  What we need now is a quick way to test things with quantal.  We can always refine on how we get the list of the releases later.
[14:58] <roaksoax> indeed
[14:59] <rvba> roaksoax: were you looking for something like this? http://paste.ubuntu.com/1187365/
[14:59] <rvba> ?
[15:00] <roaksoax> rvba: ah yes!! thanks!
[15:00] <rvba> welcome :)
[15:51] <rvba> roaksoax: I think I found why the dhcp config is not being written: see bug 1046397.
[15:53] <roaksoax> rvba: awesome!!
[15:53] <roaksoax> thanks
[16:00] <rvba> roaksoax: I'll let you triage that bug and, if you think my suggestion is right, then the fix should be pretty straightforward.
[16:07] <roaksoax> rvba: seems the right fix to me
[17:08] <roaksoax> rvba: so now the only "issue" would be how to correctly display the problem
[17:08] <roaksoax> err
[17:08] <roaksoax> the release name
[17:08] <roaksoax> as in "Precise Pangolin (12.04 - LTS)"
[17:08] <roaksoax> or similar
[17:52] <roaksoax> allenap: still around?
[18:47] <robbiew> RoAkSoAx: should a 'sudo dpkg-reconfigure maas' also restart apache2 afterwards?
[18:47] <roaksoax> robbiew: it should yes!
[18:47] <robbiew> hmm...didn't for me...but maybe it was a fluke
[18:48]  * robbiew tests again
[18:48] <roaksoax> robbiew: it won't
[18:48] <roaksoax> I removed that
[18:48] <roaksoax>  but will at it again
[18:48] <robbiew> lol
[18:49] <roaksoax> just checked :)
[18:49] <robbiew> ;)
[18:50]  * robbiew checks the box for today for "contribution to opensource"...and goes to lunch
[18:50] <robbiew> :P
[18:50] <roaksoax> hahaha
[18:50] <roaksoax> have a good one
[18:50] <robbiew> thx
[18:50] <roaksoax> cause
[18:51]  * roaksoax is still contributing :)
[18:51] <robbiew> I admire such over achievers
[18:51] <roaksoax> lol
[19:41] <Daviey> robbiew: RMS is proud of you today.
[19:50] <allenap> roaksoax: Hi there, can I help? I have just a couple of minutes though :-/
[19:55] <allenap> roaksoax: If you do need me, email and I'll try and reply this evening, or get me tomorrow before about 1630 UTC. Have a good evening!
[20:14] <roaksoax> allenap: no worries, enjoy your evening