[e2e] "PMTUD using options" draft

[=?gb2312?q?Jing=20Shen?=]

In past month we monitors access line by 'fping'
access router interface address and customer's router
interface address with 20 packet every 20 seconds. The
total amount of links monitored is about 230, and this
number will be increased soon.

The topology of our network could be graphed as:

Monitor
Host   ------L3 Switch ---- L3 Access Switch ---
Customer router

The route along the path is unique.

The monitoring result shows, the difference in RTT
between access interface and customer's router
interface may differ much(as an extrem example, 3.92ms
to access switch interface, 637.55us to customer's
router interface,100Mbps ethernet link, 4.5Mbps
maximum utilization,12hour average ). But some other
ports do not show too much difference between values
of interfaces interconnected. I think the difference
is due to ping packet processing overhead introduced
by L3 switch.

One the other hand, there also exists difference when
ping to different ethernet ports on the same access
switch, although the difference is not so much as
previous one.

Reading those post of this thread before, I think
there may be difference in manufacture's
implementation which. These difference derives to
difference in great difference between forwarding &
ping packet processing.( someone told me Cisco only
allows 5% CPU time with ICMP packet processing)





 --- Jon Crowcroft <Jon.Crowcroft at cl.cam.ac.uk>
µÄÕýÎÄ£º> sense of  humour failure detection alert
> 
> btwe my laptop can run incremental dijkstra on 
> the internet abo ut    once evey 10 sec and ita an
> old ibook
> 
> perhaps cisco should invest in apple
> In missive
>
<Pine.GSO.4.50.0402131932490.3386-100000 at argos.ee.surrey.ac.uk>,
> Lloyd Wood typed:
> 
>  >>On Fri, 13 Feb 2004, Jon Crowcroft wrote:
>  >>
>  >>> routers that send anything to a "central"
> processor are going to
>  >>> have a backplane or switch hotspot/bottleneck
> problem as well as a
>  >>> cpu problem and are out of the ark
> designs/dinosaurs for sure....
>  >>
>  >>These days, laptops are multihomed - supporting
> multiple wired and
>  >>wireless links. But everything goes to a central
> processor, and
>  >>so laptops are out-of-the-ark dinosaurs as far as
> routing goes.
>  >>
>  >>I do hope nobody is proposing routing research
> based around using
>  >>laptops for e.g. adhoc mesh communication. Or
> peer-to-peer systems
>  >>based on these obviously dinosaur general-purpose
> computers.
>  >>
>  >>L.
>  >>
> 
>
>><http://www.ee.surrey.ac.uk/Personal/L.Wood/><L.Wood at eim.surrey.ac.uk>
>  >>
> 
>  cheers
> 
>    jon
>  

=====
Jing Shen

Data Communication Center
HangZhou TeleCom 
HangZhou ZJ 310027
P.R.China

' spamcontrol '

_________________________________________________________
Do You Yahoo!? 
ÍêÈ«Ãâ·ÑµÄÑÅ»¢µçÓÊ£¬ÂíÉÏ×¢²á»ñÔù¶îÍâ60Õ×ÍøÂç´æ´¢¿Õ¼ä
http://cn.rd.yahoo.com/mail_cn/tag/?http://cn.mail.yahoo.com