=== hatchetation_ is now known as hatchetation === alexpilotti_ is now known as alexpilotti === harlowja is now known as harlowja_away === shardy_afk is now known as shardy [09:04] smoser: 0.3.2 :) [09:10] smoser: regarding r293 and r294, I can only use file injection or rebuilding cirros from source as a means to configuring the proper resizefs mode, right? [09:11] or I could use cmd execution (in this case: sed or just echo) in userdata, though that would probably come in to late for resizefs to occur on the current boot [13:46] harmw, fil injection or image modification for sure. [13:48] userdata will happen too late. [13:48] harmw, the plan would be for in 0.4 or something to have a newer kernel [13:49] and newer krenel gets magic fast online resize [13:49] so i'd enable it by default. [14:33] ah ok [14:33] fair enough [14:41] perhaps it'd be nice to have 2 prebuild images online, one with and one without resizefs [14:41] since I'm to lazy to roll my own image :p [14:41] i really dont want to complicate things. its really easy to patch an image before upload. [14:41] well, actually for cirros its more difficult isnt it :) [14:42] i'm open to suggestions on it. [14:42] the newer kernel will magically make everythign wonderful. [14:42] (i'm actually not kidding, its up to 100x faster) [14:42] true, but I can imagine that'll take atleast a year :p [14:42] for online resize. [14:43] yea you've told that bfore, sounds promissing fo [14:43] well, yeah. i know. i suck. and the other thing i want in 0.4.0 would be some udev like thing (maybe 'mdev') which is theonly way i can reasonably support running on microsoft hypervisor (name eludes me now) [14:44] hyper-v [14:44] yes, I noticed a bugreport on it [14:45] but let's just do a 0.3.2 first, ok :) [14:46] would cirros benefit from a tool to show more environment stats? hypervisor, ram, cpu and the likes? [14:47] basically extending cirros-status === harlowja_away is now known as harlowja [18:08] smoser anything u need from me releated to bug https://bugs.launchpad.net/cloud-init/+bug/1136343 ? [18:10] harlowja, nah. thats fix-released in cloud-init. [18:10] i made it public today [18:10] as juerg opened a dupe of it. [18:10] i have no intent of SRUing that to 12.04 [18:10] k [18:15] smoser: any thought about my question earlier on? [18:16] harmw for the network settings writing question, http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/view/head:/cloudinit/distros/freebsd.py#L220 [18:16] probably something simple to write, i can even do, but not sure what the freebsd network format is :_P [18:16] :-P [18:16] it's just some stuff going into /etc/rc.conf [18:17] shouldn't be that hard, indeed [18:17] harlowja, sure. extending cirros-status would be fine. and useful i think. [18:17] sounds good to me smoser [18:17] lol [18:17] lol [18:17] bah. harmw ^ [18:17] harmw, yeah, definitely 0.3.2 is not including hyperv support. [18:17] or a 3.13 kernel :) [18:18] indee [18:18] harlowja: I guess I can look into the _write_network stuff later tonight [18:18] sweet, if u want, i can try to figure it out also, but freebsd experts would probably do it quicker ;) [18:19] haha [18:19] * harmw is by no means an expert [18:19] more expert than me :-P [20:07] harlowja: openstack doesn't put any network stuff in metadata, I have to use configdrive for that right? [20:07] correct [20:07] but u can just expect that its the format at http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/view/head:/cloudinit/distros/net_util.py#L24 [20:07] if it isn't provided, then just don't write [20:08] I'm including that indeed [20:09] TypeError: _read_hostname() takes at least 2 arguments (1 given) [20:10] did I break that (freebsd, ofcourse) [20:10] did u do self._head_hostname? [20:10] self._read_hostname [20:11] this is a new checkout from launchpad, I didn't touch thta part of the code [20:11] lets dive in :) [20:12] got it [20:12] jeez [20:13] http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/view/head:/cloudinit/distros/freebsd.py#L85 [20:14] the filename parameter is not supposed to be there [20:15] hmmm, that would do it :-P [20:16] but sucks I didn't notice the error before [20:23] AHA [20:23] http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/revision/927 [20:23] smoser: you broke it :p [20:36] harlowja: https://code.launchpad.net/~harmw/cloud-init/freebsd-static-networking [20:36] woot [20:37] I havent tested that just yet though [20:37] since I'm not using configdrive atm [20:37] harmw should be easy to add a unittest that can verify what was written [20:37] that 'simulates' input and the right output [20:38] yea [20:38] could you write one? or should I do that? [20:38] if only i know what the expected output should be ;) [20:38] (I dont know how laborintensive such a thing is) [20:38] http://bazaar.launchpad.net/~harmw/cloud-init/freebsd-static-networking/view/head:/tests/unittests/test_distros/test_netconfig.py can probably be extended [20:38] ah ok [20:38] def test_simple_write_ub(self): -> def test_simple_write_bsd(self): [20:39] Ill look into that, thanks :) [20:39] np [20:39] that'd be sweet, seanwbruno if u wouldn't mind looking that over when it arrives [20:39] since u probably can verify all that stuffs :) [21:15] harlowja, what'd i break? [21:17] you introduced a parameter to a function, causing it to break once called (freebsd.py) :) [21:17] http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/view/head:/cloudinit/distros/freebsd.py#L85 [21:22] hm.. [21:22] fix? [21:22] just undo the change on that line [21:22] (dropping the filename argument) [21:25] pylint cries [21:25] ooooooh [21:26] for not a good reason. [21:26] its odd. but since it is in the parent class and even declared an abstractmethod there [21:26] we shoudl have the same [21:27] so maybe this: http://paste.ubuntu.com/7012683/ [21:27] hm yea, I was thinking the same [21:29] harmw, can you just verify that that fixses it really quick [21:29] and then i'll commmit [21:29] sorry for breaking [21:32] sure, gimme a sec [21:35] smoser: works :) [21:40] pushed [21:56] harlowja: I've extended that netconfig test a little, how should I proceed? [21:58] harmw commit it and there u go ;) [21:58] ? [21:58] oh, hm [21:58] I was expecting some of just running that test here, local [22:59] harlowja: its commited [23:00] I've pushed some other fixes in freebsd.py as wll [23:00] ah, u should just be able to run nosetests and have it locally run