After looking into it I think the root issue is that the block devices are incorrectly
named in new 2.6.40-4 kernel. I created an issue:
On Tuesday, August 09, 2011 07:52:56 AM Marek Goldmann wrote:
> Hi,
>
> In BoxGrinder we're seeing some issues when upgrading to latest kernel on
> AMI's, which was reported here:
>
>
https://issues.jboss.org/browse/BGBUILD-289
>
> You can see in that ticket full boot log (console log in AWS terms) from a
> started instance with newest kernel. Jeremy pointed out that the issue may
> be related to newest kernel: 2.6.40-4:
>
> [quote]
>
> I came across this issue searching for a possibly related problem.
>
> I had been running Fedora 15 on EC2 successfully using the 1.6 Boxgrinder
> AMI. Yesterday, I applied the latest updates, including
> kernel-2.6.40-4.fc15.x86_64 and added a boot entry in /boot/grub/menu.lst.
> On reboot, I saw similar messages in the console log to what was mentioned
> here. Attaching my boot EBS volume to another instance and changing the
> kernel back to kernel-2.6.38.8-35.fc15.x86_64 allowed my instance to boot
> normally.
>
> [/quote]
>
> Anyone can tell me more after reading the console log?
>
> TIA
>
> --Marek
looks like selinux is wanting to relabel the filesystem and that is failing.
Dennis
--
devel mailing list
devel(a)lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel