[CLOUDSTACK-2318] upgrade from 2.2.14 -> 4.1.0 : unable to start an existing running vm after stopping it manually

• Type: Bug
• Status: Closed
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 4.1.0
• Fix Version/s: 4.1.0
• Component/s: KVM
• Security Level: Public (Anyone can view this level - this is the default.)
• Labels: nfs security-groups sharedstorage
• Environment: CentOS 6.4 .. 2.6.32-358.2.1.el6.x86_64 #1 SMP Wed Mar 13 00:26:49 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
• Assignee: Wei Zhou
• Reporter: Shashi Dahal
• Created: 03/May/13 09:29
• Updated: 04/Jun/13 18:49
• Resolved: 08/May/13 16:46

Description

Hi,

I was able to successfully upgrade from 2.2.14 to 4.1.0
Successfully in the sense that the upgrade steps worked without issues.

For the VMS that were running before the upgrade, I could do the following

reboot
stop
change password - it gave a new password

I could not start an existing VM because the way the paths are in the database is different.
I had to apply this database fix to get an existing stopped VM to start state.

update volumes set path=SUBSTRING_INDEX(`path`, '/', -1);

This is discussed here: http://mail-archives.apache.org/mod_mbox/incubator-cloudstack-dev/201206.mbox/[e-mail]%3E

It would be great if this can be checked/included in the 4.1.0 release also.

Thanks,
Shashi