From: Howard C. Berkowitz (hcb@gettcomm.com)
Date: Fri Jul 04 2003 - 14:29:28 GMT-3
At 12:04 PM -0500 7/4/03, James.Jackson@broadwing.com wrote:
>Agreed, if stress testing you'll want to use 64 byte packets in order to
>maximize pps before you hit bandwidth limitations. Similarly it's fair to
>say that figures you see stated for pps were achieved with small packets but
>this does not imply a direct correlation between packet size and pps
>performance. If on the other hand you're trying to simulate typical internet
>traffic you can use the rough 7:4:1 distribution for 64 byte, 512 byte, 1500
>byte respectively.
>
>Cheers,
>James
Do look at http://www.isi.edu/in-notes/rfc2544.txt for standard
benchmark methodology.
As an aside, but appropriate for this being the CCIE list, it would
strike me that performing and explaining a benchmark would be a very
reasonable test requirement. Might be a case study on the written.
This archive was generated by hypermail 2.1.4 : Wed Aug 06 2003 - 06:52:24 GMT-3