volume attachment fails on a paused instance

[Yogev Rabl]

Description of problem:
When running volume attachment on a paused instance, the process seem to succeed, the volume is registered as 'in-use' and there's no error in the nova compute log.
But when resuming the instance the OS doesn't see the newly attached volume and when trying to detach the volume it fails with the error "DiskNotFound: No disk at vdb" in the nova compute log.

Version-Release number of selected component (if applicable):
python-nova-2013.2.3-6.el6ost.noarch
openstack-nova-compute-2013.2.3-6.el6ost.noarch
python-novaclient-2.15.0-4.el6ost.noarch
openstack-nova-common-2013.2.3-6.el6ost.noarch

How reproducible:
100%

Steps to Reproduce:
1. Launch an instance with a working OS with a QCOW image.
2. Pause the instance.
3. Attach a volume to the instance.
4. Resume the instance.
5. In the instance: run the command:
# lsblk

Actual results:
The instance's OS doesn't see the attached volume.

Expected results:
The instance's OS should see the volume.

Additional info:
The nova-compute log is attached. (please notice that the log is very long,

[Thang Pham (thang-pham)]

I believe this is a duplicate of bug #1242366. Although the steps to reproduce the bug is slightly different, I believe it is the same bug.

[Yogev Rabl (yrabl)]

Yes, it does seem like the same core problem. I'll add my scenario to the bug.
Thanks.