Reset from image failures

Submitted by Lucian on Fri, 03/11/2011 - 06:21

Hi,

I have a test Centos guest which I tried to "reset from image" to Debian Lenny. It fails. See fail1.txt attachment.

I then created a separate Lenny guest which I tried to reset to Centos, this also failed, see fail2.txt.

If in both attempts I choose to reset to the same OS as the original install then all is fine...

Suggestions?

Status: 
Closed (fixed)

Comments

Submitted by JamieCameron on Fri, 03/11/2011 - 12:17

Odd, I haven't seen this one.

Were both systems created from and reset from Cloudmin-provided images?

Submitted by JamieCameron on Fri, 03/11/2011 - 13:26

Wait, I see the cause of this now - it only happens when a VM is reset using a different image, or more specifically an image with a different partition table.

I will fix this in Cloudmin 5.4.

Submitted by Lucian on Fri, 03/11/2011 - 15:59

Ok, great!

Btw, any reason to not just reset the partition table as well? Also, is there any real need for a /boot partition? Seems more like complicating things but I have a feeling I'm missing something :)

Submitted by JamieCameron on Fri, 03/11/2011 - 16:17

So the issue is that Cloudmin caches for each system the offset of the root partition into the disk image, in bytes. During a reset to a new image this may change .. but the cached offset wasn't being cleared.

Submitted by JamieCameron on Fri, 03/11/2011 - 16:19

So the issue is that Cloudmin caches for each system the offset of the root partition into the disk image, in bytes. During a reset to a new image this may change .. but the cached offset wasn't being cleared.

Submitted by JamieCameron on Fri, 03/11/2011 - 16:25

So the issue is that Cloudmin caches for each system the offset of the root partition into the disk image, in bytes. During a reset to a new image this may change .. but the cached offset wasn't being cleared.

Submitted by Issues on Fri, 03/25/2011 - 18:21

Automatically closed -- issue fixed for 2 weeks with no activity.