[Starcluster] create_image.py

[Dan Yamins]

> 1) Launch an instance of the starcluster AMI
> 2) run create_image.py and ostensibly generate an exact copy of the
> original AMI
> 3) verify that the new AMI exists and is in a new S3 bucket
> 4) shut down the original instance
> 5) launch a new instance based on the "new" AMI
> 6) Observe that the starcluster -s script take a reaaaaaaaaly long time to
> start up
> 7) Log in to instance and observe that SGE is not running
> _____________
>


Mark, thanks for your emails.   I too have had problems creating re-bundled
AMIs , in just the manner you mention.   That is, after rebundling and
trying to start and instance from the new AMI, the start-up script gets
stuck in the SGE-installation stage, and the SGE binaries are never running.


However, when I followed the directions for rebundling an AMI directly from
Amazon's AWS site (
http://docs.amazonwebservices.com/AWSEC2/latest/GettingStartedGuide/creating-an-image.html)
 I was able to create very stable working AMIs from the starcluster base
images (both 32bit and 64 bit).   Have you tried these directions?    If
not, they might work better that create_image.py.

Overall, bundling is a somewhat tricky process and I have had much worse
problems rebundling functional AMIs from *other* sources.

It would be great though, if the problem that you mentioned could be figured
out, because I think it is something that a number of users have
experienced.

Dan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/starcluster/attachments/20100108/f9b4ed56/attachment.htm