[02:42] <bigjools> smoser: is there anything documenting how to get cloud-init to run scripts sent in user data?
[02:43] <smoser> https://help.ubuntu.com/community/CloudInit
[02:43] <smoser> simplist thing is if it starts with '#!' it gets executed rc.local time.
[02:43] <bigjools> great
[02:44] <bigjools> thanks
[06:15] <bigjools> jam, no luck with hangouts today!
[06:15] <bigjools> read HACKING.txt in the tree, it explains exactly how to set up
[06:16] <bigjools> <bigjools> jam, no luck with hangouts today!
[06:16] <bigjools> <bigjools> read HACKING.txt in the tree, it explains exactly how to set up
[06:16] <jam1> k
=== jam1 is now known as jam
[06:21] <jam1> bigjools: so clearly I was wrong. As it just died without any connection running...
[06:21] <bigjools> :)
[06:21] <jam1> but I think we got through the bulk of it
[06:21] <bigjools> yeah
=== jam1 is now known as jam
[06:21] <bigjools> the next step is for you to play with the code
[06:23] <jam> bigjools: making sure I get it now. What is the review/landing process? Similar to launchpad?
[06:23] <jam> Do you use the same bot?
[06:24] <bigjools> jam: I have tarmac running, so we just set MPs approved and it lands them
[06:24] <bigjools> same review process, same code standards
[06:24] <jam> bigjools: it is running on your machine(s)?
[06:24] <bigjools> on devpad .... not ideal but works
[06:25] <bigjools> we also have a jenkins instance that runs tests when it sees new revisions
[06:32] <jam> bigjools: for more integration-level testing? is it a different set of tests?
[06:32] <bigjools> jam: diogo is working on that
[06:33] <jam> ah, so that is the automation of package installs, etc?
[06:33] <bigjools> yep
[09:21] <czajkowski> roaksoax: ello remember I asked about http://askubuntu.com/questions/178088/configuring-for-automatic-boot-over-avahi-kernel-requires-an-x86-64-but-i686-det  any way we can let the user know if this is possible or something that may happen so we can at least answer it
[11:24] <roaksoax> czajkowski: hi! will do
[11:24] <roaksoax> jtv: howdy!
[11:30] <czajkowski> roaksoax: thanks
[12:30] <roaksoax> smoser: do you think maas-import-ephemerals has something to do with wrongly permissions being set when running maas-import-pxe-files
[12:30] <roaksoax> that is if you have looked at it recently
[12:31] <smoser> what is the issue?
[12:31] <melmoth> hey, i have a strange feeling that my manual change in /etc/nova/nova.conf have been overriden at a certain stage (they are back to what i set in the nova-controller charm)
[12:32] <melmoth> was i tired, or is that  what actually happen ? And if so, how can i reconfigure an alread deployed charm to have new settings ?
[12:32] <roaksoax> smoser: so on maas-import-pxe-files /var/lib/tftpboot/maas/amd64/generic/precise/commissioning ends un havin 0600 permission, when it needs 0755
[12:32] <roaksoax> smoser: however, /var/lib/tftpboot/maas/amd64/generic/precise/install does have the same permissions. The only "apparent" different thing is the fact
[12:32] <roaksoax> that the ephemeral images are download with mas-import-ephemerals
[12:33] <smoser> roaksoax, its a directory with 0600?
[12:33] <roaksoax> smoser: yes
[12:34] <roaksoax> permissions 600
[12:34] <smoser> jtv, ping
[12:35] <smoser> it looks like install_tftp_image in maas-import-ephemerals leaves a tempdir around
[12:36] <smoser> i really wish we had useful comments in commits
[12:43] <roaksoax> hehe :)
[12:55] <roaksoax> smoser: were you able to find the reason behind the permissions issue? or was it because of a tempdir?
[12:57] <smoser> roaksoax, i was just looking, found that i didn't have a good commit message, got off on a side track.
[12:57] <smoser> will continue looking now.
[12:58] <smoser> roaksoax, the 0600 really confuses me
[12:58] <smoser> (on a directory)
[12:58] <smoser> but it wouldnt surprise me if
[12:58] <roaksoax> smoser: err sorry, it is 0700
[12:58] <smoser> ok
[12:58] <smoser> then yeah, i suspect
[12:58] <smoser>  maas install_pxe_image
[12:58] <smoser> is copying permissions of
[12:58] <smoser> --image="$tmpdir"
[12:58] <smoser> and tmpdir was created with mktemp
[12:58] <smoser> (so it will have 0700)
[12:59] <roaksoax> indeed, that might actually be the case
[12:59] <roaksoax> since making the directory with permissions specified, is no different
[12:59] <roaksoax> so it must be it then
[13:00] <smoser> where is 'install_pxe_image'? in maas?
[13:00] <smoser> i dont see it
[13:00] <roaksoax> smoser: src/provisioningserver/pxe/install_image.py
[13:01] <roaksoax> that's the file that does that
[13:02] <smoser> allenap, have you tested vdenv after your "Change tabs to spaces and remove trailing whitespace." (revno: 830)
[13:02] <smoser> it seems to me that you  might have broken it
[13:03] <smoser> vdenv/zimmer-build/ud-build.txt had tabs that were not indentation
[13:06] <smoser> oh shoot. thats just wrong.
[13:07] <smoser> jtv didn't leave a temp file around, but only because 'install_image' removes the directory its given
[13:07] <smoser> ie:
[13:07] <smoser>  maas-provision install-pxe-image --image=/
[13:07] <smoser> ==
[13:07] <smoser>  rm -Rf /
[13:08] <smoser> jtv, and then also your install_tftp_image does 'umask a+r' without returning the umask.
[14:21] <flacoste> matsubara: I've created https://launchpad.net/~maas-maintainers/+archive/daily-qa-ok
[14:21] <flacoste> matsubara: that's where you should copy the packages when they pass the QA tests
[14:25] <matsubara> thanks flacoste
[15:46] <allenap> smoser: No, I didn't run it again, so I'm sorry if that broke it. But... sounds like it needs some tests ;)
[15:47] <smoser> allenap, i'm not certiain that dpkg preseed requires tab delimited input
[15:47] <smoser> but i think it does
[15:47] <smoser> and you removed those tabs.
[15:55] <allenap>  /o\
[16:40] <smoser> hey all.
[16:40] <smoser> anyone want to test:
[16:40] <smoser> https://maas.ubuntu.com/images/ephemeral/daily/precise/20120825/
[16:41] <smoser> i'd appreciate input
[17:29] <roaksoax> smoser: where you able to look at the permissions issue? I did this: http://paste.ubuntu.com/1174233/ but seems kinda ugly
[17:31] <smoser> i didn't look at it at all, no.
[17:32] <smoser> bu what you ahve doesn't look terrible to me.
[17:38] <roaksoax> smoser: alright, I guess that's better than nothing
[17:39]  * roaksoax lunch
[18:43] <jgcampbell300> can anyone tell me where the dns configs happen with maas ?
[19:04] <jgcampbell300> ok ... i checked .. /etc/dnsmasq.conf and it says addn-hosts = /var/lib/cobbler/cobbler_hosts ... so i check that file and its empty .... i wonder if i have finaly found the place i can add in dns name servers
[19:30] <roaksoax> smoser: did you, by any chance, updated the ephemeral image today?
[19:30] <smoser> roaksoax, sort of.
[19:31] <roaksoax> smoser: uhmm it is failing to enlist
[19:31] <smoser> there is a daily, and it was accidently published as released for a couple hours
[19:31] <roaksoax> ah I see that miht be the one
[19:32] <roaksoax> i'll update the images
[19:33] <smoser> but i'm interested in knowing about the failure, because it should not have failed
[19:35] <roaksoax> smoser: it was about not being able to access arhicve.ubuntu.com
[19:36] <smoser> you mean import-ephemerals was failing ?
[19:36] <roaksoax> smoser: err nope, I'm sorry
[19:37] <roaksoax> s/ephemeral/cloud-init
[19:37]  * roaksoax is in slow motion today
[19:39] <smoser> roaksoax, well, i intended to put it as 'daily' for testing
[19:39] <smoser> so this is part of what i'd hoped to shake out
[19:39] <roaksoax> alright
[19:39] <smoser> and i was trying to test also
[19:40] <roaksoax> smoser: weird!! i imported new images and now everything works like a chaarm
[19:40] <smoser> roaksoax, well, that woudl just indicate a regression in the images that i crated
[19:41] <roaksoax> smoser: right but it was coud-init failing to work
[19:41] <roaksoax> err
[19:41] <roaksoax> or access the archives
[19:42] <smoser> do you have the console log?
[19:42] <roaksoax> smoser: nope unfortunately, erased everything
=== cinerama_ is now known as cinerama