Help Interpreting traceroute output

[Jerry Feldman]

On Monday 18 October 2004 1:28 pm, dsr at tao.merseine.nu wrote:

> Performance testing is best done between the points in question
> with the protocol in question using a payload as similar to
> "typical" as possible.
I agree with DSR's assessment. Both ping and traceroute are excellent tools 
and may be indicators of network problems, but there are some tools you can 
use that may give a better indication of true bandwidth measurement. 
Remember that when you ping and traceroute you are also adding DNS into the 
mix. In the past, I have used tools like ttcp. But there are other network 
performance tools available. There is also some differences between TCP and 
UDP. Once you establish a tcp connection, since you have established a 
virtual circuit, your packets will generally travel through the same 
routers where every UDP (and ICMP for that matter) could take a different 
path. (Note that TCP has the capability for each packet to travel via a 
different route, but in reality it will be a virtual circuit).
-- 
Jerry Feldman <gerald.feldman at hp.com>
Partner Technology Access Center (contractor) (PTAC-MA)
Hewlett-Packard Co.
550 King Street LKG2a-X2
Littleton, Ma. 01460
 (978)506-5243