Some of you may remember about half a year - year ago, I asked for feedback (and got some) for an internet draft on MTU issues in the tunnels which are set up in the network. This is a topic with pretty high operational relevance.

I've revised the draft again based on some really good feedback, and it's getting very near publication soon.

So, I'd like to solicit more comments (on- or off-list). Please comment within about a week or so.


    Tunneling techniques such as IP-in-IP when deployed in the middle of
    the network, typically between routers, have certain issues regarding
    how large packets can be handled: whether such packets would be
    fragmented and reassembled (and how), whether Path MTU Discovery
    would be used, or how this scenario could be operationally avoided.
    This memo justifies why this is a common, non-trivial problem, and
    goes on to describe the different solutions and their characteristics
    at some length.