[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: A RED problem
Kartik,
It's true that for any one drop, it shouldn't matter if the packet dropped
is at the tail of the queue or elesewhere within the queue. However, the
choice of which packet to drop will affect later packet transmissions.
Dropping one packet rather than another can cause a flow to backoff that
otherwise wouldn't, thus affecting future packet transmissions and drops.
At least, that's my understanding; although I've never actually played with
the drop_tail_ setting.
Jeremy Ethridge,
Computing Technology Lab,
Nortel Networks.
> -----Original Message-----
> From: Kartik Sudeep [SMTP:[email protected]]
> Sent: Tuesday, July 13, 1999 10:33 AM
> To: [email protected]
> Subject: A RED problem
>
> I wanted to know if the "drop_tail_" parameter for RED in ns gives rise to
>
> different number of dropped packets depending on whether it is set to
> "true"
> or "false".
>
> According to me, drop_tail_ set to true only means that the last packet
> has
> to be dropped. And setting it to false implies that any packet in the
> queue
> can be dropped.
> => The number of dropped packets should be the same in either case.
>
> Please correct me if I am mistaken.
>
> I have tried to run a test-suite for RED, but obtained different results
> for
> the two cases. So I am a bit confused!
> (try running test-suite-sack.tcl sackB4 with drop_tail_ true and false).
>
> Thanx,
> Kartik
>
>
> ______________________________________________________
> Get Your Private, Free Email at http://www.hotmail.com