[03:20] <ThiagoCMC> What is this on my maas syslog: dhcpd: Can't create new lease file: Permission denied ?
[04:28] <shang> anyone seen this issue before? http://pastebin.ubuntu.com/1296760/
[04:29] <ThiagoCMC> MaaS or Juju related?!
[04:30] <ThiagoCMC> I'm working only with MaaS about a week now and, I am very disappointed... I found at least 5 BUGs.
[04:31] <ThiagoCMC> And I'm still did not touch Juju... I'm seeing that I'll find more and more BUG as I'm going deeper...
[04:31] <ThiagoCMC> Damn...   :-/
[04:32] <shang> are they stopping you from deploying workload via juju?
[04:33] <shang> I found few issues, but most of them have workaround
[04:33] <ThiagoCMC> I still didn't touch Juju...
[04:33] <ThiagoCMC> Since I'm finding too much bugs with MaaS...
[04:34] <shang> were you using 12.04? i.e. the maas with cobbler?
[04:34] <ThiagoCMC> As follows: BUGs: 1069535, 1069584, 1069570, 1069603 and 1067929...
[04:34] <ThiagoCMC> Ubuntu 12.10
[04:34] <ThiagoCMC> damn bot...
[04:35] <ThiagoCMC> lol
[04:36] <ThiagoCMC> See!? I am very disappointed with MaaS and with Canonical... Those guys release a "stable" system that simple do not work out-of-the-box, as they say.
[04:37] <ThiagoCMC> A week of work just to figure out the MaaS problems...
[04:37] <ThiagoCMC> (facepalm)
[09:56] <melmoth> is it me or is the swift-proxy charm not installing swift at all ?
[09:56] <melmoth> http://paste.openstack.org/show/21968/
[09:56] <melmoth> (that s the content of the charm.log on the node where i deployed swift-proxy)
[11:28] <mgz> rvba: what's the right way of generating html in django outside a template? question is related to doing escaping correctly.
[11:32] <rvba> mgz: there is a tool in django: django.utils.html.escape (https://docs.djangoproject.com/en/dev/ref/utils/#module-django.utils.html)
[11:32] <mgz> thanks!
[11:32] <rvba> np
[13:26] <melmoth> anyone deploying swift with keystone ?
[13:27] <melmoth> i m confused about who is suppose to create the service user ? me, the swift charm ?
[13:28] <melmoth> if it s swift-proxy, i do not find any trace of a keystone stuff in the install hooks (plus it mention
[13:28] <melmoth> " until we use keystone or another real auth system" ....
[14:52] <rbasak> I assumed that bug 1069850 was intentional?
[14:55] <torment> running maas in 12.10 - can i just edit /etc/maas/dhcpd to put in the right next-server ip? or should i be doing this in the database
[14:56] <matsubara> torment, you can run dpkg-reconfigure maas-region-controller
[14:56] <matsubara> if you edit manually, you also need to sudo service maas-dhcp-server restart
[14:56] <torment> k let me try
[14:58] <torment> hmm, this restarts apache on the private ip
[15:02] <torment> oh nop i didnt wait long enough ;)
[15:33] <bigjools> mgz: around?
[15:33] <bigjools> or jam?
[15:33] <jam> bigjools: I'm around
[15:34] <bigjools> jam: are you free to come round to the dining area?
[15:34] <jam> In the main hall? Sure
[15:34] <bigjools> easier to talk directly :)
[15:58] <mgz> bigjools: I'm only around virtually till tomorrow evening
[15:58] <bigjools> mgz: no worries, jam helped
[15:59] <jam> mgz: someone is running maas, and ran into the bug with parsing memory, which landed as rev 1266, but it turns out Q has r1264
[15:59] <jam> apparently the SRU is waiting on someone to approve it.
[15:59] <jam> mgz: as an aside, it seems that if we fail to detect memory (setting it to 0) you can't bootstrap to the node, because the juju provider requests some memory
[15:59] <jam> (I don't know what the default constraint is, but it doesn't match 0)
[16:43] <mgz> bug 1064638 does note the workaround for bootstrapping when memory is not recorded
[17:31] <torment> "The region controller does not know whether any boot images have been imported yet."
[17:31] <torment> i ran maas-import-pxe-images
[17:31] <torment> nothing particular in the logs
[17:45] <torment> bug 1067929 oh i see
[20:10] <ThiagoCMC> torment, I made a small procedure to not hit that "5 minutes" bug... It is there on launchpad...
[20:20] <torment> ThiagoCMC: hey, i saw that... i really dont want to reinstall again tho
[20:20] <ThiagoCMC> I reainstall it more than 20 times....
[20:20] <ThiagoCMC> lol
[20:20] <torment> i got all my nodes added, they all say ready, but i can't juju bootstrap at this point
[20:21] <torment> im not sure if it has to do with the error still being at the top
[20:21] <torment> ERROR No matching node is available.
[20:21] <ThiagoCMC> Honestly, I give it up... I'll not use MaaS + Juju until next 04/2013... It is unstable (at least for me)...
[20:22] <torment> yeah, its coming along nicely from last release
[20:23] <ThiagoCMC> I would LOVE to use MaaS + Juju, it is a very nice idea and have a great future... But it does not work for me... One week later... I'm still facing problems and problems...
[20:23] <ThiagoCMC> :(
[20:49] <matsubara> torment, the no matching node is available might be due to juju constraints
[20:50] <matsubara> torment, what's the command you're running?
[20:50] <matsubara> torment, and there's a script by smoser to mirror the ephemeral images
[20:51] <matsubara> torment, you could use that while testing so you don't need to re-download all the images again from the internet
[20:51] <matsubara> you can mirror them locally and point your import-pxe-files to it
[20:53] <matsubara> torment, this is the mirror tool: https://code.launchpad.net/~smoser/+junk/mirror-query
[21:22] <torment> ok i will use that next time
[21:23] <torment> is there a fix for the 5 minute image check to get rid of the message?
[23:04] <torment> heh, i deleted all the interfaces from the cluster controller, and theres no way to add any
[23:08] <torment> is there an easy way to reinstall just maas without reloading the OS?