[ovs-dev] OVN/OVS Split amended proposal

[Ben Pfaff]

On Mon, Nov 05, 2018 at 12:49:31PM -0500, Mark Michelson wrote:
> On 11/02/2018 03:59 PM, Ben Pfaff wrote:
> >On Fri, Nov 02, 2018 at 03:13:35PM -0400, Mark Michelson wrote:
> >>With this in mind, we'd like to propose an amended plan.
> >
> >[...]
> >
> >>In order to facilitate this, OVS will need to be outfitted with ways for OVN
> >>to probe for the existence of features at runtime. The most obvious use of
> >>this would be ovn-controller querying ovs-vswitchd about the existence of
> >>certain OpenFlow match fields and actions.
> >>
> >>We believe this is a more realistically achievable goal, is less likely to
> >>conflict with other development projects, and is less likely to result in
> >>bugs being introduced. Since this plan directly implements phases that are
> >>necessary for the full separation, it acts as a stepping stone rather than a
> >>band-aid or stop-gap.
> >>
> >>What are your thoughts on this?
> >
> >Seems reasonable to me.
> >
> >My only real comment is that I don't think new OVS features are needed
> >to query for match fields and actions.  I think that we can do that with
> >what's already available.
> >
> 
> That's great news.
> 
> Can you think of any other runtime incompatibilities that might exist
> between OVN and OVS? I had considered differences in the running OVS
> database. But I also don't think that will require anything new at runtime
> to support.

The database shouldn't require extra runtime support because the client
library is already careful to probe for the newer server features that
it wants and fall back to older features if necessary.

If I think of anything else, I'll bring it up.