[ubuntu-cloud] deboostrapped lucid image from maverick fails on ec2 (with and without AKI/custom kernel)

[Scott Moser]

On Sun, 21 Nov 2010, Spike Spiegel wrote:

> Hey,
>
> finally some real progress.
>
> On Sun, Nov 21, 2010 at 11:51 AM, Spike Spiegel <fsmlab at gmail.com> wrote:
>
> > that's all. in the meantime I'm gonna try to get one of those images,
> > strip it out of stuff and then rebundle it and see what happens if I
> > pick an amazon AKI.
>
> this worked (can ssh in and all) and answered some of the questions
> above. I've downloaded the lucid uec tar.gz, chrooted and customized
> it to my likings, including removing all the kernels to reduce the
> size as much as possible while growing the img size to use the whole
> 10GB root of a small ec2 instance (otherwise you only get the 1.5 the
> img is created as).

correct.  We resize the image up to 10G before uploading.  The lower size
is useful for UEC where the default disk is smaller, and also reduces
download size and space on your disk if you aren't using the full 10G.

>  As an AKI I've actually used the one reported
> here[1] (aki-6603f70f). I used the default ramdisk ARI, altho I'd very
> much like to understand what that implies since that image is not
> supposed to use a ramdisk afaik and I'm not sure which ones do and why
> and how you'd pick an ARI if needed. Anyway, I'm a step closer.

Do not specify an ARI.  When you bundle, or register, do not specify
'--ramdisk'.  There is, unfortunately, no way to explicitly specify
'--ramdisk NONE'.

Thats the right kernel to be using now.

> Next I'll try to rebuilt the whole thing with vmbuilder, especially
> since the current lucid kernel does not seem to have bridging support
> available which I need for lxc, so guess I'll have to work out the
> pv-grub stuff after all and run my own kernel.

Really?  Please open a bug if its missing function that it needs.  I
assume you're referring to -ec2, but the use case is quite valid, so we
can likely get the -ec2 kernel in Lucid serviced with new modules.