[ovs-dev] [PATCH] Rename "xs-vif-mac" external_ids key to "peer-mac", for generality.

Ben Pfaff blp at nicira.com
Wed Sep 1 23:50:29 UTC 2010


On Wed, Sep 01, 2010 at 03:58:08PM -0700, Ben Pfaff wrote:
> On Wed, Sep 01, 2010 at 03:43:11PM -0700, Jesse Gross wrote:
> > On Wed, Sep 1, 2010 at 3:13 PM, Jeremy Stribling <strib at nicira.com> wrote:
> > >> On Wed, Sep 01, 2010 at 02:15:09PM -0700, Jesse Gross wrote:
> > >>> I think this a good idea in theory but the name and description
> > >>> initially made me think that this was for PPP or something.  I know
> > >>> that VIF is somewhat Xen centric but I think the name and description
> > >>> were better as they were before, just without the xs prefix.  In
> > >>> practice, I'm pretty certain that this field will always be used for
> > >>> hypervisors, so I'm not sure there is much value in making it more
> > >>> generic beyond that.
> > > Actually, Justin approved "peer-mac" on purpose instead of "vif-mac" (though
> > > I don't think he's online to defend himself).  I think the reasoning is that
> > > you'll want to be able to label MACs in the physical world, where VIFs don't
> > > make any sense.
> > 
> > The issue is that in neither case is it describing the address of a
> > peer.  A true peer would probably be either another switch in the
> > physical world or a tunnel endpoint in the virtual one, neither of
> > which has a MAC that is of interest to us.
> > 
> > What about "remote-mac" or "attached-mac"?  The description as a point
> > to point link also seems really strange to me.  Why not simply say
> > "the MAC address of the device or virtual machine attached to this
> > interface"?
> 
> I care this much >< about the name.

After discussion we agreed on "attached-mac" and I pushed it that way.




More information about the dev mailing list