Re: types of traffic in tcp congest control
J Wu <jinw@comp.leeds.ac.uk> Sun, 12 May 2002 23:26 UTC
X-Authentication-Warning: cslin135.leeds.ac.uk: jinw owned process doing -bs
Date: Mon, 13 May 2002 00:26:43 +0100 (BST)
From: J Wu <jinw@comp.leeds.ac.uk>
X-X-Sender: <jinw@cslin135.leeds.ac.uk>
To: =?ISO-8859-1?B?tPPA7g==?= <liy666@sina.com>
Cc: <tcp-impl@lerc.nasa.gov>
Subject: Re: types of traffic in tcp congest control
In-Reply-To: <20020512135045.28137.qmail@sina.com>
Message-ID: <Pine.LNX.4.33.0205122354460.30118-100000@cslin135.leeds.ac.uk>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=ISO-8859-1
X-MIME-Autoconverted: from 8bit to quoted-printable by cslin.leeds.ac.uk id g4CNQiet013582
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by lombok-fi.lerc.nasa.gov id UAA13109
Sender: owner-tcp-impl@grc.nasa.gov
Precedence: bulk
Status: RO
Content-Length: 2076
Lines: 51
On Sun, 12 May 2002, ´óÀî wrote:
> Hello,
>
> I need some help with the following :
>
> i had read a paper writen by Steven H.low, Internet Congestion Control ,published by ieee control systems magazine , which describes the types of traffic as follows:
>
> most TCP connections are "mice" (short, but requiring low latency), but a few long TCP connections ("elephants," which can tolerate latency) generate most of the traffic. By controlling the network around a state with full queues, the elephants subject the mice to unnecessary loss and queueing delays...
>
> i don't think that "mice" or "elephants" have direct relationship with the long or short latency, the length of latency may be caused by the type of application, such as voice, multimedia, as so no.
>
> Please check this viewpoint, thanks.
>
> Y. Lee
>
> Yong Lee, Ph.D Candidate
> Network Lab.
> Dept. of Computer Science & Engineering
> Southeast University
> Sipailou2#(PC.210096)
> Nanjing, P.R.China
> (Phone) +86-25-3791010
>
> ===================================================================
> ÐÂÀËÃâ·Ñµç×ÓÓÊÏä (http://mail.sina.com.cn)
> ÐÂÀË·ÖÀàÐÅÏ¢£º¶þÊÖÊг¡×ßÒ»×ߣ¬¸Ã³öÊÖʱ¾Í³öÊÖ£¡ (http://classad.sina.com.cn/2shou/)
>
Hi Li,
I think that the "length of latancy" in your email is representing for
network delays, which caused by queuing delay at routers and propagation
delay of the medias. So, this kind of lantancy is absolutly decided by the
transmittion subnet and have nothing to do with the upper layer
applications.
As my understanding of S.Low's paper that the mice stands for small scale
transmition with delay sensitive and the elephant stands for large scale
transmition. The small scale transmitions will not cause any congestion,
but the large scale transmitions will. If cut off the large scale
transmitions for a litte to eliminate the congestion, which can benifit
a lot for the delay sensitive small scale transmitions.Also, as the
"elephant" is huge that harms little for gives way to "mice".
--
Jin Wu
School of Computing,
University of Leeds.
Tel: +44 113 2336806
- Re: types of traffic in tcp congest control J Wu
- Re: types of traffic in tcp congest control Joe Touch
- Re: types of traffic in tcp congest control J Wu
- Re: types of traffic in tcp congest control Joe Touch
- RE: types of traffic in tcp congest control Douglas Otis
- Re: types of traffic in tcp congest control J Wu
- RE: types of traffic in tcp congest control J Wu
- Re: types of traffic in tcp congest control Joe Touch
- Re: types of traffic in tcp congest control Alan Cox
- Re: types of traffic in tcp congest control Alan Cox
- Re: types of traffic in tcp congest control Alan Cox
- Re: types of traffic in tcp congest control Joe Touch
- Re: types of traffic in tcp congest control Anumita Biswas
- Re: types of traffic in tcp congest control Joe Touch
- Re: types of traffic in tcp congest control Anumita Biswas
- Re: types of traffic in tcp congest control Joe Touch
- Re: types of traffic in tcp congest control Alan Cox
- RE: types of traffic in tcp congest control Douglas Otis