Great Circle Associates Firewalls
(April 1995)
 

Indexed By Date: [Previous] [Next] Indexed By Thread: [Previous] [Next]

Subject: Firewall Performance Limitations
From: padgett @ tccslr . dnet . mmc . com (A. Padgett Peterson, P.E. Information Security)
Date: Mon, 24 Apr 95 14:28:59 -0400
To: "firewalls @ greatcircle . com"@UVS1.dnet.mmc.com

Well theoretically, the SOL translates to about 1 ns/foot. In "real world"
terms I always allow 1.5 so 6000 miles would be about 50 ms. Now given that 
the "sunk" time is about 15 ms (what the RTT is to a machine across the room)
you have about 65 ms for send-transit-bounce-transmit-receive. Now for
me to ping the last stop (5 nodes from here) before leaving Orlando is 50 ms.
The next node in Denver averages 110 ms so for Orlando - Denver you spend
about the same amount of time routing as you do travelling. 

Will say that the SEA-Clark-Chicksands RTT in 1968 was measured in 
whole seconds for 100 wpm TTY comm.

Considering the amount of processing you can do in a single millisecond, I
suspect that so long as the 'wall is not busy with something else, the 
processing time will be lost in the noise.

						Warmly,
							Padgett

Indexed By Date Previous: unsubscribing
From: craig @ actech . com (Craig Mills)
Next: Looking for recommended conferences
From: Tom Ceresini <Tom . Ceresini @ shrmed . com>
Indexed By Thread Previous: Re: firewall performance limitations
From: George Mullins <george @ wicked . neato . org>
Next: Re: firewall performance limitations (fwd)
From: fc @ all . net (Dr. Frederick B. Cohen)

Google
 
Search Internet Search www.greatcircle.com