[openstack-dev] [nova] [api] [placement] strategy for placement api structure

Jay Pipes jaypipes at gmail.com
Mon Jun 20 14:25:27 UTC 2016

On 06/20/2016 10:21 AM, Sylvain Bauza wrote:
> Le 20/06/2016 16:04, Jay Pipes a écrit :
>> On 06/17/2016 11:23 AM, Sylvain Bauza wrote:
>>> +1000 yes to that. Now the devil could be in the details, in particular
>>> how we implement the WSGI server, the corresponding WSGI app and the
>>> associated routing, and that's exactly the problem below.
>> We shouldn't be implementing a WSGI server *at all*. The fact that one
>> cannot run Nova inside a true WSGI server (i.e. Apache/mod_wsgi,
>> uwsgi, etc) is a bad thing.
> Okay, fair point :-)
> That still leaves the routing question up, tho :-)
>>> I certainly understand the concerns of adding yet another library. To be
>>> honest, I tend to even agree with the statement that we could possibly
>>> use Routes without explicitly use nova.wsgi, right ?
>>> In the review, you explain why you don't trust Routes and I respect
>>> that. That said, are those issues logged as real problems for our API
>>> consumers, which are mostly client libraries that we own and other
>>> projects we know, like Horizon ?
>>> If that is a problem for those, is there something we could improve,
>>> instead of just getting rid of it ?
>> For the record, I'm very much in favor of the approach Chris has taken
>> in a framework-less implementation with nothing other than the (tiny)
>> selector library as a dependency. I'd like to see the work move forward.
> Okay, my open question (because I'm not expert in that) was rather to
> understand why Routes was not possible to be something usable for the
> new placement API, if that's something we use in Nova too.

Routes would indeed be possible to use. And I wouldn't have an issue 
using Routes, though I prefer selector due to its simplicity and 
documentation readability and that controllers are specified as actual 
callables, not strings.


More information about the OpenStack-dev mailing list