[09:45] <blaisebool> hello, how can I run an ansible playbook bootstrapped with cloud-init ?
[13:16] <smoser> rharper, hm.
[13:16] <smoser> rharper, bah. i must have branched ubuntu/devel. will re-do from master.
[13:21] <smoser> rharper, https://code.launchpad.net/~smoser/cloud-init/+git/cloud-init/+merge/304415 looks saner now.
[13:23] <blaisebool> hello, how can I run an ansible playbook bootstrapped with cloud-init ?
[13:25] <smoser> blaisebool, cloud-init doesn't have any specific support for ansible.
[13:25] <smoser> but a simple '#!' as user-data that does whatever you would normally do should work for you.
[13:25] <smoser> cpaelzer, https://bugs.launchpad.net/cloud-init/+bug/1618572
[13:25] <smoser> i can't reproduce either. i had a simplified version also.
[13:26] <smoser> i suspect it might be related to overlayroot
[13:26] <cpaelzer> hmm
[13:26] <cpaelzer> ah you are here
[13:26] <cpaelzer> only saw in #server before
[13:27] <smoser> so maas ephemeral environment is running overlayroot
[13:27] <cpaelzer> roaksoax not here
[13:27] <cpaelzer> :-/
[13:27] <smoser> /etc/apt/.... thats on overlayroot
[13:27] <smoser> i'll try on a serverstack
[13:27] <smoser> and let you know how i did it.
[13:28] <roaksoax> cpaelzer: here
[13:28] <cpaelzer> roaksoax: hi
[13:28] <cpaelzer> roaksoax: just relaized discussing with smoser that you were missing here
[13:29] <cpaelzer> roaksoax: smoser is assuming it might be related to overlayroot and tries to recreate on serverstack
[13:29] <roaksoax> cpaelzer: ack!
[13:29] <roaksoax> cpaelzer: i'm setting up a different environment too
[13:29] <roaksoax> cpaelzer: so you can access there
[13:31] <smoser> i suspect this is overlayroot as the cause.
[13:31] <smoser> we've seen surprisinggly little issues with overlayroot, but it does have some wierdities
[13:32] <roaksoax> smoser: but wouldn't that also affect the chroot ?
[13:32] <roaksoax> smoser: but wouldn't that also affect the chroot for curtin ?
[13:33] <smoser> roaksoax, no. because the chroot is not on an overlay
[13:41] <smoser> this is overlayroot related, almost sure.
[13:41] <smoser> i just did this:
[13:41] <smoser> a.) launch instance in openstack (serverstack) of yakkety
[13:41] <smoser> b.) echo "overlayroot=tmpfs" | sudo tee /etc/overlayroot.local.conf
[13:41] <smoser> c.) sudo reboot
[13:41] <smoser> console log: http://paste.ubuntu.com/23116093/
[13:41] <smoser> horay!
[13:42] <smoser> seems like kernel regression in overlayroot to me
[13:43] <roaksoax> fun!
[13:48] <smoser> ok. suck.
[14:06] <cpaelzer> grmbl
[14:07] <cpaelzer> why can't there ever be a day where things just work :-)
[14:07] <cpaelzer> smoser: since you call it a regression - is the same working for xenial or before?
[14:07] <cpaelzer> smoser: I don't know if you had the time
[14:30] <smoser> so what i know so far
[14:30] <smoser> trusty , i can do
[14:30] <smoser>  boot image
[14:30] <smoser>  sudo sh -c 'echo overlayroot=tmpfs > /etc/overlayroot.local.conf && reboot'
[14:30] <smoser> everything fine
[14:30] <roaksoax> smoser: btw.. on that note, any ETA on getting cloud-init to xenial (with the apt support ?)
[14:30] <smoser> that seems to stop working at wily (systemd)
[14:30] <smoser> roaksoax, well, why would i SRU that when it would break you
[14:31] <smoser> (as you shown)
[14:31] <smoser> :)
[14:31] <roaksoax> smoser: but if you say that's a kernel regression :) then that at least would get into -proposed
[14:31] <roaksoax> smoser: so it is ready to land once kernel fixes the kernel ?
[14:31] <smoser> well, theres stills ome things i do not understand there.
[14:31] <roaksoax> smoser: oh I see!
[14:31] <smoser> i dont think cloud-init is at fault
[14:32] <smoser> but we may end up needing to make changes
[14:32] <roaksoax> smoser: yeah doesn't seem to be that case
[14:41] <roaksoax> right