Cell Relay Archive

[Thomas Frieling <thomas.frieling@arcormail.de>]

Dominic Richens schrieb:
> 
> Daniel Horner <dnh@nortelnetworks.com> wrote in message
> yvgg7l96k03d.fsf@nortelnetworks.com">news:yvgg7l96k03d.fsf@nortelnetworks.com...
> > I'm wondering about the simultaneous application of EPD and PPD. Is
> > there a benefit to using both of these discard policies at once? If I
> > understand correctly, then PPD would still catch the tail-end of
> > frames that "snuck past" the EPD threshold. Ie. the beginning of the
> > AAL5 frame was transmitted when the queue was just under the EPD
> > threshold.
> 
> PPD is there to discard all remaining cells in a frame, regardless of why a
> cell was discarded. EPD only ever discards the first cell in a frame, if it
> is received when the queue fill is above the EPD discard threshold (for
> instance 80% of the queue size).  Hence PPD is needed to "clean up" after
> EPD has discarded the first cell in a frame.

Really? And I mean to remember that Marconi ASX-Switches act like
this:
Early Packet Discard begins to discard cells when the buffer
situation is not too tight yet. The switch decides to discard an
entire packet and lets the one before pass untouched. All cells
that form the following packet are discarded and not just the
first!
Partial Packet Discard happens when the buffer fills up rapidly.
The switch cannot wait until a new packet arrives. It has to
start discarding cells immediately. This means, that some cells
of a packet or frame are already transmitted. To avoid confusion
at the receiver the last cell of the frame may be transmitted.
This is also described here:
http://cbel.dcrt.nih.gov/~jelson/ip-atm/node21.html
The author of that page refers to:
A. Romanow and S. Floyd
Dynamics of TCP traffic over ATM networks. 
IEEE Journal on Selected Areas in Communications, May 1995. 
Available on the Internet at
ftp://ftp.ee.lbl.gov/papers/tcp_atm.ps. 


Thomas Frieling