Re: No WRED for 3560 , why not ?

From: johngibson1541@yahoo.com
Date: Sat May 12 2007 - 21:37:25 ART

• Next message: johngibson1541@yahoo.com: "Care about where the bottle neck of traffic is ?"
• Previous message: johngibson1541@yahoo.com: "No WRED for 3560 , why not ?"
• Maybe in reply to: johngibson1541@yahoo.com: "No WRED for 3560 , why not ?"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

My reasoning does not make sense.

If 3550 is not hitting the thresholds, 3550's WRED
doesn't take effect. At this situation, if 3560
hits threshold and if 3560 is the bottle neck
of the path, the TCP sessions contending the same
queue will get global synchronization symptom .

I am curious about few vendors give situations
when the location of the bottle neck matters.
Everyone seems to be satisfied with controlling
a particular device's QOS behavior.

John

>Because the random dropping at 1 point in a traffic
>path is enough ? Rely on some router or the access
>layer 3550 switch to do random dropping ?
>
>Multiple random dropping points is not better than
>just 1 random dropping point ? If so, why ?

• Next message: johngibson1541@yahoo.com: "Care about where the bottle neck of traffic is ?"
• Previous message: johngibson1541@yahoo.com: "No WRED for 3560 , why not ?"
• Maybe in reply to: johngibson1541@yahoo.com: "No WRED for 3560 , why not ?"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

This archive was generated by hypermail 2.1.4 : Fri Jun 01 2007 - 06:55:20 ART