RE: RoE header and mapper thoughts: packet sizes
Hi Jouni,
I'm not sure that we can recommend on the packet size at that time. First let's see what is the mapping and what are the mandatories fields that we need, then we can state what is the frame size, based on that work.
The frame size will be output from the mapping...
Rgrds
--Raz
-----Original Message-----
From: stds-1904-3-tf@xxxxxxxx [mailto:stds-1904-3-tf@xxxxxxxx] On Behalf Of Marek Hajduczenia
Sent: Saturday, February 21, 2015 2:46 AM
To: 'Jouni Korhonen'; STDS-1904-3-TF@xxxxxxxxxxxxxxxxx
Subject: RE: RoE header and mapper thoughts: packet sizes
Jouni,
It ought to be up to implementation to pick what frame sizes it can generate. I believe applications (mapper in this case) would try to optimize performance and hence aim for larger frames. From standard perspective, it is not something I believe needs to be specified in the standard.
Product-wise - certainly, but that is not something that ought to be placed in te standard to guarantee interop
Marek
-----Original Message-----
From: stds-1904-3-tf@xxxxxxxx [mailto:stds-1904-3-tf@xxxxxxxx] On Behalf Of Jouni Korhonen
Sent: Friday, February 20, 2015 5:09 PM
To: STDS-1904-3-TF@xxxxxxxxxxxxxxxxx
Subject: RoE header and mapper thoughts: packet sizes
Folks,
Do we want to state something about the recommended packet sizes? I am rather stubborn on the minimal RoE header size (I still want to believe 32bits is enough). How about then the payload size? Any opinions there e.g.
trying to stay below 512 bytes (which would be more than 512 on wire)?
- Jouni
--
Jouni Korhonen, Ph.D, Associate Technical Director CTO Office, Networking, Broadcom Corporation
O: +1-408-922-8135, M: +1-408-391-7160