operational: icmp echo out of control?

Tue May 28 17:59:54 UTC 2002

Chris Woodfield wrote:
 >...the next-best method is to probe HTTP ports, but we don't want to
> have to pull down thousands of web pages just to get performance stats.

Why not just passively measure the time it takes to send actual
traffic to actual clients?   It shouldn't take too much talent
to use the aggregate data for an evolving best guess for each client 
request.

KL