RE: Requirements for IP-in-IP Tunnel MTU Assurance

["Templin, Fred L" <Fred.L.Templin@boeing.com>]

Fred, 

Cutting down some on text from the earlier discussions:

> Why is it proper to fragment at a tunnel endpoint and not  
> in the general case?

Because the tunnel endpoint is the head-end of a link over
which an assured MTU has to be presented to layer 3. This
falls within the realm of layer 2 link adaptation; not
network-based fragmentation.

> Why would this be preferable to finding the path MTU?

Tunnel MTU assurance is not a replacement for finding the
path MTU; the tunnel is just an ordinary link on the path
and the tunnel endpoint would participate in classic path
MTU discovery the same as for any other link.

Fred
fred.l.templin@boeing.com