[Rdo-list] [Openstack] rdo havana to icehouse: instances stuck in 'resized or migrated'

[Nikola Đipanov]

On 05/20/2014 07:06 PM, Julie Pichon wrote:
> On 20/05/14 17:08, Dimitri Maziuk wrote:
>> On 05/20/2014 03:59 AM, Julie Pichon wrote:
>>> On 19/05/14 18:14, Dimitri Maziuk wrote:
>>>> On 05/19/2014 11:15 AM, Julie Pichon wrote:
>>>>>
>>>>> I had a chat with a Nova developer who pointed me to the following patch
>>>>> at https://review.openstack.org/#/c/84755/ , recently merged in Havana
>>>>> and included in the latest RDO Havana packages. Resize specifically is
>>>>> one of the actions affected by this bug, you might want to check that
>>>>> you're running the latest packages on your Havana node(s) and see if
>>>>> this might help to resolve the problem?
>>>>
>>>> [root at irukandji ~]# yum up
>>>> ...
>>>> No Packages marked for Update
>>>>
>>>> So -- no, that's not it, unless the patch hasn't made it into the rpms yet.
>>>
>>> Could you provide the version numbers for the openstack-nova packages?
>>
>> Icehouse node:
>>
>> [root at squid ~]# rpm -q -a | grep openstack
>> openstack-selinux-0.1.3-2.el6ost.noarch
>> openstack-nova-api-2014.1-2.el6.noarch
>> openstack-utils-2014.1-1.el6.noarch
>> openstack-nova-compute-2014.1-2.el6.noarch
>> openstack-nova-network-2014.1-2.el6.noarch
>> openstack-nova-common-2014.1-2.el6.noarch
>>
>> Havana node:
>>
>> [root at irukandji ~]# rpm -q -a | grep openstack
>> openstack-nova-common-2013.2.3-1.el6.noarch
>> openstack-nova-network-2013.2.3-1.el6.noarch
>> openstack-nova-api-2013.2.3-1.el6.noarch
>> openstack-selinux-0.1.3-2.el6ost.noarch
>> openstack-nova-compute-2013.2.3-1.el6.noarch
>> openstack-utils-2013.2-2.el6.noarch
>>
>> Both nodes are
>>
>> [root at squid ~]# cat /etc/redhat-release
>> CentOS release 6.5 (Final)
>>
> 
> Thanks for the information. I'm adding my colleague Nikola to this
> thread, he's familiar with Nova and should be better able to help.
> 
> Julie
> 

Hi Dimitri,

So for this kind of upgrade to work, you will need to set an RPC version
cap so that all your icehouse computes know they are talking to possibly
lower version nodes and downgrade their messages.

This can be done (as described in much more detail on [1]) using the
[upgrade_levels] section of your nova.conf file.

Please let me know if this works and if you need further assistance.

Thanks,

Nikola Đipanov, SSE - OpenStack @ Red Hat

[1] http://openstack.redhat.com/Upgrading_RDO_To_Icehouse