[openstack-dev] 7/17 state of the gate (you know, fires)

Matt Riedemann mriedem at linux.vnet.ibm.com
Sat Jul 18 15:07:57 UTC 2015



On 7/18/2015 9:31 AM, Matt Riedemann wrote:
>
>
> On 7/18/2015 8:54 AM, Matt Riedemann wrote:
>>
>>
>> On 7/17/2015 7:51 PM, Matt Riedemann wrote:
>>> We started the day with a mock 1.1.4 release breaking unit tests for a
>>> few projects (nova, cinder, ironic at least).
>>>
>>> The nova blocker was tracked with bug 1475661.
>>>
>>> We got a g-r change up to block mock!=1.1.4 but that didn't fix the
>>> issue because oslo.versionedobjects pulls mock in before pip processes
>>> the !=.  Luckily lifeless reverted the regression and released mock 1.2
>>> so we should be OK on that front for now.
>>>
>>> However, cinder/glance/nova, which gate on the ceph job, are blocked by
>>> bug 1475811 which is a regression with the rbd driver in glance_store
>>> 0.7.0, which was put into upper-constraints.txt today.
>>>
>>> I have a patch up for glance_store here:
>>>
>>> https://review.openstack.org/#/c/203294/
>>>
>>> And an exclusion of 0.7.0 in g-r here:
>>>
>>> https://review.openstack.org/#/c/203295/
>>>
>>> If the glance core team could get around to approving that fix and
>>> releasing 0.7.1 it would unblock the gate for glance/cinder/nova and
>>> make me happy.
>>>
>>
>> The g-r block on glance_store 0.7.0 is approved but can't merge because
>> one of the jobs keeps failing to build qpid-proton:
>>
>> http://logs.openstack.org/95/203295/1/check/gate-requirements-integration-dsvm/10073dd/console.html#_2015-07-18_13_16_32_766
>>
>>
>>
>> So if anyone from oslo.messaging or qpid can help with that it'd be good.
>>
>
> Got a change up here for devstack:
>
> https://review.openstack.org/#/c/203354/
>

devstack was the wrong place, we found the regression from yesterday, 
and have a revert posted:

https://review.openstack.org/#/c/203355/

-- 

Thanks,

Matt Riedemann




More information about the OpenStack-dev mailing list