From: P729 (p729@cox.net)
Date: Tue Mar 25 2003 - 19:46:56 GMT-3
Indeed, if you believe that round-trip delay is relevant as I once used to.
I recently sat in a QoS bootcamp given by the TAC and the instructor made a
very compelling argument (based on internals, as in how the queue scheduler
operates) for selecting the Bc so that you get the time interval you want
(quantum).
In some of the CCO examples, you may see a quantum of, say, 125 mS. This
could theorectically use up a lot of VoIP delay budget because even if a
small voice packet can get clocked out in 25 mS, nothing else will happen
for another 100 mS. I think the smallest quantum is 10 mS.
I asked, "but what about the 1.5 sec. round-trip delay, etc." and the
instructor said something like, "those docs really need to be updated."
Regards,
Mas Kato
https://ecardfile.com/id/mkato
----- Original Message -----
From: "Scott M. Livingston" <scottl@sprinthosting.net>
To: "'eric'" <namaste@pacbell.net>; "'Ccielab'" <ccielab@groupstudy.com>
Sent: Saturday, March 22, 2003 7:39 PM
Subject: RE: Car Formula
Eric,
Your first formula is correct!
Per Cisco and others;
http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122cgcr/
fqos_c/fqcprt4/qcfpolsh.htm#1000920
"We recommend the following values for the normal and extended burst
parameters:
normal burst = configured rate * (1 byte)/(8 bits) * 1.5 seconds
extended burst = 2 * normal burst"
-scott
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
eric
Sent: Saturday, March 22, 2003 8:15 PM
To: Ccielab
Subject: Car Formula
Can anyone confirm the Car formula
I do believe its one of these just not sure which.
Bc = cir/8 x 1.5
Be = Bc x 2
CIR 64000
Bc 12000
Be 24000
or is it
Bc = cir/8 x .5
Be = Bc x 2
CIR 64000
Bc 4000
Be 8000
This archive was generated by hypermail 2.1.4 : Sat Apr 05 2003 - 08:51:44 GMT-3