[00:30] <SpamapS> anybody have a good guide for using lxc with libvirt?
[16:11] <koolhead11> hi kim0
[16:11] <kim0> koolhead11: hey :)
[16:12] <koolhead11> howdy?
[16:12] <kim0> koolhead11: very well .. how about yourself
[16:12] <koolhead11> all well thanks!!
[16:12] <kim0> Awesome
[16:13] <koolhead11> did you posted the new blog?
[16:17] <kim0> koolhead11: which one
[16:17]  * kim0 brb in 5 mins
[19:32] <SpamapS> smoser: On natty, can I use cloud-init with a local file for the metadata?
[19:40] <smoser> yes
[19:41] <SpamapS> smoser: awesome thanks..............................
[19:43] <SpamapS> smoser: so, dump the data into /var/lib/cloud/seed/nocloud  ?
[19:43] <smoser> the current trunk doc is right.
[19:43] <smoser> http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/files/head:/doc/examples/seed/
[19:44] <smoser> 'nocloud-net' rather than 'nocloud' will guarantee that network is up.
[19:44] <smoser> you care about that if you do any boothooks
[19:51] <Kyle__> Has anyone here installed images from http://uec-images.ubuntu.com on your UEC cloud?
[20:09] <Kyle__> I ask because each time I try to run one, they don't start.
[20:56] <obino> Kyle__: did you manage to start other images? is this a problem only with images form the URL you sent?
[20:58] <Kyle__> obino: Only with the images from that site.  The ones I download from links the extras page all work.
[20:59] <obino> they stay pending?
[20:59] <obino> the ones that doesn't work I mean :)
[21:00] <Kyle__> The ones that don't work are all from that site, yes.
[21:02] <obino> which kernel/initrd are you using?
[21:02] <Kyle__> The ones bundled with those images.
[21:02] <obino> or do you see any error in nc.log?
[21:03] <obino> nc.log may contains some errors
[21:03] <Kyle__> Opening it back up, didn't see any errors last time.
[21:03] <obino> if not, one option would be to try to start them by hand
[21:03] <obino> if there is no error, eucalyptus (uec) handled the image to libvirt to start it
[21:04] <obino> you can look at /var/log/libvirt for logs about it
[21:04] <Kyle__> How would you do that (I have a way I've tried, but don't want to color your thoughts with a prehaps bad way I've been trying to use).
[21:04] <obino> or you can try to run it by hand to check what the proble it
[21:04] <obino> in eucalyptus.conf on the NC there is an option to have MANUAL_CLEANUP
[21:04] <obino> set it, and restart the NC
[21:04] <Kyle__> MANUAL_CLEANUP?  WHat will that do?
[21:04] <obino> after that the failed instance files will be left behind
[21:05] <obino> you will see them in $INSTANCE_PATH
[21:05] <obino> search for a directory with the name of your instance id
[21:05] <obino> inside you will have everything you need to start it with virsh
[21:05] <Kyle__> Cool.  Thank you.
[21:05] <obino> you will need to modify the *xml file
[21:05] <Kyle__> It needs virsh, not just kvm/qemu-kvm?
[21:06] <obino> network will not work as is for example
[21:06] <obino> and the paths may be different
[21:06] <Kyle__> K.
[21:06] <obino> virsh (libvirt) will call kvm/qemu the same way eucalyptus does
[21:06] <Kyle__> Hum.  In nc.log I see walrus_request(): due to error, removing /var/lib/eucalyptus/instances//admin/i-48C707B9/kernel
[21:06] <Kyle__> error: failed to download or prepare into /var/lib/eucalyptus/instances//admin/i-48C707B9/kernel
[21:06] <Kyle__> Failed to prepare images for instance i-48C707B9 (error=1)
[21:06] <Kyle__> Humm.
[21:07] <obino> ok, then you don't even get the files on the NC
[21:07] <obino> do you happen to donwloads the admin credentials multiple times?
[21:08] <Kyle__> Err I think I've done it more than once... why?
[21:10] <obino> there is known issue about that
[21:10] <Kyle__> So.... what can be done?
[21:11] <obino> you should use only the last one downloaded to bundle the image
[21:11] <obino> the bundling will encrypt the image with the cert
[21:12] <obino> you should use the last one downloaded to do that