[ovs-dev] [PATCH] [Patch V2] OVN: Logical-Physical Separation and L2 SW Gateways

[Russell Bryant]

On Mon, Mar 21, 2016 at 6:07 PM, Darrell Ball <dball at vmware.com> wrote:

>
>
>
>
>
> On 3/21/16, 9:41 AM, "dev on behalf of Russell Bryant" <
> dev-bounces at openvswitch.org on behalf of russell at ovn.org> wrote:
>
> >On Mon, Mar 21, 2016 at 8:47 AM, Darrell Ball <dlu998 at gmail.com> wrote:
> >
>
> >
> >On a related topic, I have a need for a localnet port that exists on a
> >subset of hypervisors.  In this case, I still only want to model it with a
> >single logical localnet port.
> >
> >That would work today as long as no VIF ports are created on a chassis
> that
> >does not have that network available locally.  We need to improve this on
> >the OpenStack integration side to make sure that we never try to bind a
> >port to a chassis where that network is not available.  I was originally
> >thinking that this work might help solve that problem, but now I'm not
> sure
> >that's true.  Using the physical endpoints table would require a logical
> >port and a physical endpoint for every hypervisor attached to that
> network,
> >which could be hundreds.
> >
> >I think this can all still work, but for this use case, we just need tom
> >store the bridge mappings in the southbound database, as well.  I was
> >thinking of storing it as an external_id on the Chassis record.
> >
> >How does that sound?
>
>
> If physical endpoints are used to check the chassis accessibility to a
> given network,
> multiple physical endpoints would be needed as you described.
>
>
> A provider network access is usually specified at a chassis/physical port
> level
> (plus the encap part).
> The existence of a physical access bridge comes from the port connectivity
> knowledge.
>
>
> With physical endpoints usage, by binding of the localnet port, the
> connectivity is
> being defined directly and physical access bridges can be auto created as
> needed based
> on this configuration.
>
> Bridge-mapping requires some preprocessing behind the scenes to find the
> physical port mapping
> on a given chassis to a network and then create another bridge if needed
> and then create the bridge mapping,
> which is then checked later on port binding time to see if the chassis has
> access (i.e a bridge) to that network.
> It would seem like the physical endpt creation and binding is a more
> direct way of achieving the same result.
> However, maybe I don’t understand some requirements/constraints for
> localnet.
>

I'm not opposed to tracking the network availability per chassis info in
the southbound db.  That sounds good.

However, I'd like to avoid having to create N logical ports, where N is the
number of hypervisors attached to a network.  I want to continue to create
only 1 logical port just like today.

-- 
Russell Bryant