[Rdo-list] [rhos-list] instances with ephemeral disk can't start ?

[Pádraig Brady]

On 01/24/2014 10:43 PM, Perry Myers wrote:
> On 01/24/2014 05:16 PM, Xin Zhao wrote:
>> Hello,
>>
>> In our grizzly openstack testbed, instances that use ephemeral disk fail
>> to run, in the scheduler log there is the following errors:

>> **kwargs)\n', u"/TypeError: _create_ephemeral() got an unexpected
>> keyword argument 'max_size'/\n"]
>>
>> I see a bug report like this :
>> https://bugzilla.redhat.com/show_bug.cgi?id=1044562, I wonder if this
>> bug is fixed in grizzly or not?
> 
> This bug is currently being QA'd for an async release of RHOS 3.0/Grizzly.
> 
> It is not yet released for RHOS 3.0 since it is still in testing.  It
> should be updated soon on our content mirrors (Red Hat Network)
> 
> If it is not a candidate for the stable branch of Grizzly, then it
> wouldn't be in RDO Grizzly as RDO follows the upstream stable branches.
> 
> Vladan, since you are working on this, did it get proposed for the
> stable branch or is it RHOS only backport?
> 
>> I am using grizzly rpms from the RDO repo:
>>
>> $> rpm -qf
>> /usr/lib/python2.6/site-packages/nova/virt/libvirt/imagebackend.py
>> python-nova-2013.1.4-1.el6.noarch

Note python-nova-2013.1.4-2.el6 should be the first version with
this issue. Anyway...

> Since this is RDO related, I'm going to move this thread over to rdo-list.

This fix in already on upstream stable:
  https://review.openstack.org/#/c/57821/
and was already ported to RDO grizzly in version python-nova-2013.1.4-5.el6
This unfortunately wasn't placed in testing and in the meantime 2013.1.4-6
became available (which contains an additional CVE fix).
So I've synced 2013.1.4-6 to the RDO repos now.

thanks,
Pádraig.