[openstack-dev] [Massively distributed] Optimizing the interaction of OpenStack components

Matt Riedemann mriedemos at gmail.com
Wed May 24 14:59:24 UTC 2017


On 5/11/2017 1:44 PM, Georg Kunz wrote:
> Nevertheless, one concrete thing which came to my mind, is this proposed 
> improvement of the interaction between Nova and Neutron:
> 
> https://review.openstack.org/#/c/390513/
> 
> In a nutshell, the idea is that Neutron adds more information to a port 
> object so that Nova does not need to make multiple calls to Neutron to 
> collect all required networking information. It seems to have stalled 
> for the time being, but bringing forward the edge computing use case 
> might increase the interest again.
> 

Yes we've needed a sort of bulk query capability with the networking API 
for years. I started going down a path the other night of trying to see 
if we could bulk query floating IPs when building the internal instance 
network info cache [1] but it looks like that's not supported. The REST 
API docs for Neutron say that you can't OR filter query parameters 
together, but at the time looking at the code it seemed like it might be 
possible.

Chris Friesen from Wind River has also been looking at some of this 
lately, see [2].

But getting people looking at doing performance profiling at scale and 
then identifying the major pain points would be really really helpful 
for the upstream development team that don't have access to those types 
of resources. Then we could prioritize investigating ways to fix those 
issues to improve performance.

[1] https://review.openstack.org/#/c/465792/
[2] http://lists.openstack.org/pipermail/openstack-dev/2017-May/117096.html

-- 

Thanks,

Matt



More information about the OpenStack-dev mailing list