The IP over ATM Mailing List Archive by date

[Tim Dwight <0006078043@mcimail.com>]

I'm afraid I took this thread off on a tangent that, while not uninteresting,
is not the group's primary focus.  Let's see if I can undo this, by
stating my query more clearly.

Jon's original posting, included a statement that correlated the IP
MTU with a proper setting of the ATM "maximum burst size" (MBS) leaky
bucket parameter.

I believe the sense of this is that while over a long time scale an IP
session's rate may be substantially less than line rate, the cells which
make up an IP packet will tend to arrive back to back.  Thus the
connection's "instantaneous rate" will exceed its negotiated "average rate"
(in ATM-ese, its sustainable cell rate or SCR), invoking the policing
function.  The policing function will allow the connection to exceed
its SCR, for a period derived from MBS.

I believe that Jon suggests setting MBS = (number of cells in IP MTU).
I'm trying to figure out what this accomplishes.

Certainly one would not want MBS < MTU, since otherwise cells could
be lost long before congestion occurred simply due to the clumping
of cells from 1 packet.

But what specific good things happen if MBS = (number of cells in IP
MTU), and why?

Tim Dwight
MCI