[Thincrust-devel] Thincrust AMIs

[Bryan Kearney]

Yes.. pull down the latest adk. You can specify an appliance as follows:

test:
     kickstart: "test.ks"
     memory: "512"
     cpus: "1"
     version: "1"
     checksum: True
     s3bucket: mahBucket
     ec2kernel: testkernel
     ec2ramdisk: testrd

The ex2kerle and ramdisk only effect the ec2bundle. Is there something 
else it should hit?

-- bk

Joey Boggs wrote:
> This is due to a kernel/modules mismatch, we install the modules from 
> the 2.6.18.7-2 kernel from Fedora 8 during the conversion process since 
> it's the only available Fedora kernel. In order to do this when bundling 
> you have to specify the matching aki/ari images. This was missed during 
> bundling. Is there something in the adk configuration so far where we 
> can specify these entries? It will need to be done on all appliances ran 
> though it for them to work properly. I'm rebundling the image now and 
> will send out another email when it's available.
> 
> 
> 
> Bryan Kearney wrote:
>> This was bounced over the wekeend. I am posting it here. Joey or 
>> David, have you seen this:
>>
>>
>> <post>
>>
>> Attempted to extend and rebundle the thincrust F10 ami (ami-3e56b257) 
>> .. but it failed when doing ec2-bundle-vol.
>>
>>  mount: could not find any free loop device
>> ERROR: execution failed: "mount -o loop 
>> /mnt/my-thincrust-f10-base-20081221 /mnt/img-mnt"
>>
>> losetup -a
>>
>> returned nothing.
>>
>> I modified modprobe.conf with
>> options loop max_loop=8
>>
>> which I don't think should be required. When doing
>> modprobe loop
>>
>> got
>> FATAL: Could not load /lib/modules/2.6.16-xenU/modules.dep: No such 
>> file or directory
>>
>> So, I'm presuming I can't rebundle the F10 image due to either some 
>> issue with the F10 image or when Amazon is injecting the kernel and 
>> libraries.
>>
>> jon
>>
>>
> 
> _______________________________________________
> Thincrust-devel mailing list
> Thincrust-devel at redhat.com
> https://www.redhat.com/mailman/listinfo/thincrust-devel