The IP over ATM Mailing List Archive by date

[Craig Partridge <craig@aland.bbn.com>]

Earl:

While I agree with some of the thrust of your note, there are spots where
I think it takes a wrong turn.

    Then we found that VCs are not cheap and use resources to setup and maintain
    at all points
    in the network. So now we are looking for techniques to reduce VC useage.
    For one we start talking about making all switches packet aware so that we
    can avoid excessive VC usage. 

Yes, but we had to do that anyway once we starting allowing cells to get
dropped.  AAL 3/4 is little better than AAL5 in conserving line bandwidth
when cells get dropped.  Cell dropping forced making switches packet aware.
Multicast just adds to the pain.  [As an aside, the multicasting issues
all suggest a MID would be nice, but every time I look at the issue it
is clear that AAL 3/4 wasn't the right way to go about it and it isn't clear
we can fit a big enough MID into a 48-byte payload without killing throughput.]

    As I recall, the original plan was that network center switches could be
    inexpensive due to
    the low resource requirements - no significant buffering is required if
    proper management is 
    used for the allocation of VCCs to links and everyone behaves themselves or
    we police them
    into shape to enforce the contracts.

Yes, and a lot of us thought that idea was bogus from the start.  It only
works if you assume traffic is very nicely distributed and multiplexes
nicely.  And we knew that data wasn't going to fit that mold from the start.
The pain was that until about 1993 (when the self-similar traffic work came
along), we didn't have a good way to convincingly explain this problem to
folks who wanted to believe burst-Poisson models were perfect.

Craig