> > The format of the label is less interesting to me, but using
 > > <interface-name>/<index> seems like a fine labeling scheme to me.  Using
 > > <interface-name>:<index> would be a wink to the existing design, has
 > > that been considered? ;-)
 > 
 > yes, it was considered, and we ran away from it, because the kernel
 > also uses that name-space. What if "bge0:4"
 > mapped to the logical interface bge0:19 (or in the case of V6, 
 > bge0:19, bge0:20, bge0:21)?

Right.  It would also cause massive confusion -- e.g., ifconfig and ipadm
may show things with the same name that refer to totally different objects.

 > Already the overlap between hostname and labels is causing so much churn!
 > 
 > If I had to choose, I would prefer the interface/number constraint.
 > 
 > > Another idea would be to have a label namespace that is flexible (not
 > > strictly constrained to an interface name and index), while
 > > automatically generated labels could use a convention using an interface
 > > name and index.
 > 
 > that was where I was coming from. But seems like the hostname namespace
 > has to be kept disntinct from this space.

"Has to be kept distinct" because of my objections, or because you also
agree that it would cause confusion?

-- 
meem

Reply via email to