[openstack-dev] [OSLO] Current DB status
gkotton at redhat.com
Mon Jul 1 16:50:53 UTC 2013
On 07/01/2013 07:35 PM, Mark McLoughlin wrote:
> On Mon, 2013-07-01 at 19:33 +0300, Gary Kotton wrote:
>> On 07/01/2013 07:07 PM, Mark McLoughlin wrote:
>>> On Mon, 2013-07-01 at 18:59 +0300, Gary Kotton wrote:
>>>> On 07/01/2013 06:49 PM, Mark McLoughlin wrote:
>>>>> On Mon, 2013-07-01 at 18:32 +0300, Gary Kotton wrote:
>>>>>> On 07/01/2013 06:13 PM, Mark McLoughlin wrote:
>>>>>>> (Oslo is not an acronym)
>>>>>> can someone please clarify what the database status is in Oslo?
>>>>> What apart from the oslo.config issue do you want clarified?
>>>> Boris had concerns about the slave database support. He mentioned that
>>>> the blueprint may be blocked -
>>> It's blocked in the sense that Nova can't yet use oslo.config because of
>>> the aforementioned issue.
>>>> This patch
>>>> updates to oslo.config-1.2.0a2 - so it may address the blueprints
>>>> blocking item.
>>>> My concern here is about the process. I was under the assumption, and
>>>> may be wrong here, is that if code is approved in Oslo the is can be
>>>> consumed by other projects. Over the last few months we have had a large
>>>> amount of database issues in Neutron and moved to the Oslo DB code to
>>>> address these.
>>> Yes, we've had (and continue to have) problems - no project has been
>>> able to use new oslo.config features because of a series of python
>>> packaging related issues.
>>> If you're suggesting that e.g. the DB code in oslo-incubator should
>>> avoid requiring oslo-config-1.2.0, that wouldn't help quantum much since
>>> the dependency on 1.2.0 was added in order to enable behaviour that
>>> quantum required.
>> The Oslo DB code requires oslo.config-1.2.0a2. This config version has
>> supported for 'deprecated_opts' - which enabled the common DB code to
>> support Neutron and Nova DB configuration variables.
>> I was not aware that that the slave blueprint was blocked because it was
>> waiting for a specific config version (hopefully the Nova DB patch deals
>> with as it uses the required config). My concern was that Oslo code was
>> approved in one project (Neutron) and in another it raises question
>> (Nova). Basically it would be nice to know if there are issues with the
>> Oslo DB support and if so what needs to be done to address them.
> The blueprint says:
> "This is blocked until we can get oslo.config 1.2.0a2 into nova.
> Currently it breaks unit tests when you run with tox."
> Which is the same issue as the one described here:
More information about the OpenStack-dev