On 02/23/2015 04:07 PM, Kyle Mestery wrote: > On Sat, Feb 21, 2015 at 6:02 AM, Russell Bryant <rbry...@redhat.com > <mailto:rbry...@redhat.com>> wrote: > > On 02/20/2015 09:51 PM, Kyle Mestery wrote: > > On Fri, Feb 20, 2015 at 7:15 PM, Russell Bryant <rbry...@redhat.com > <mailto:rbry...@redhat.com> > > <mailto:rbry...@redhat.com <mailto:rbry...@redhat.com>>> wrote: > > *** Create a Neutron ML2 mechanism driver that implements the > mappings > > on Neutron resource requests > > > > Do we want to do an ML2 MechanismDriver or do a full-fledged plugin? > > Since we're starting from scratch here, we should evaluate the pros and > > cons of this decision. It does save some code, but do we envision > > running OVN with other ML2 MechanismDrivers at the same time? > > > > Note I'm not necessarily saying I'm in favor of doing a full plugin, > > only that we should think about it a bit before deciding which path to > take. > > I don't have a strong opinion here. I just thought a ML2 mech driver > was the preferred method these days. > > It is. The only reason I brought this up is that we're likely to discuss > the future of the default ML2+OVS reference implementation in Vancouver > and it's future. One possibility is spinning it all out into stackforge. > That would affect things here, so just wanted to make sure you were > aware of this. Overall, even if we start with ML2, we can quite easily > move to a standalone plugin if needed, so I think starting with ML2 > should be fine.
OK, sounds good. I think most of the important ML2 driver code would be reusable if we need to change. Thanks for the heads up. -- Russell Bryant _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
