RE: [bmwg] Meeting Minutes Review: IPsec Terminology

Excellent.  Can we agree then that the frame/packet sizes should be defined,
with an escape clause for those who want to define their own mix.  BUT
including a mandate that all tested sizes be reported, ensuring
apples-to-apples.

Thoughts?

-----Original Message-----
From: Merike Kaeo [mailto:kaeo@merike.com]
Sent: Tuesday, October 28, 2003 3:21 PM
To: sporetsky@quarrytech.com; sporetsky@quarrytech.com;
brian.talbert@mci.com; mbustos@ixiacom.com; bmwg@ietf.org
Cc: ipsec-term@external.cisco.com
Subject: RE: [bmwg] Meeting Minutes Review: IPsec Terminology

>We are going to end up debating Packet Sizes for an "Imix" instead of IPsec
>benchmarking.  I would add to your list 40 bytes, 48 bytes, and 512 bytes
>packets.  40 and 48 bytes are common POS benchmarks and 512 bytes is a
>common packet size on the Internet.
>
>The Imix definition is not unique to IPsec.  It applies to many BMWG drafts
>and RFCs.  One alternative is to reference the phrase "a mix of packet
>sizes" in the IPsec draft and have it _defined_ in the General Benchmarking
>Terminology draft that Al and Kevin plan to write.

ok....I agree having  the mix defined generally is a better solution 
overall.  I would hate for the benchmarking docs for IPsec get hung up on 
test packet size debates :)  One would hope that people are getting smarter 
and asking for what packet sizes were tested when looking at perfomance 
data and comparing numbers.......

- merike

_______________________________________________
bmwg mailing list
bmwg@ietf.org
https://www1.ietf.org/mailman/listinfo/bmwg