Re: Jumbo Frames in 10GbE?

["Henry Ngai" <hngai@xxxxxxxxxxxx>]

I would have to say no to even consider Jumbo frame. Please consider the
following questions:

1. How do we bridge over to existing system without fragmentation?

2. If we do fragmentation, how to regroup them?

3. FDDI to Ethernet back to FDDI is a good example of a nightmare regarding
fragmentation.

4. What happen to Jabber function implemented by existing physical layers in
exiting 802.3? They will chop off your large frame without fragmentation.

5. If we are to do fragmentation, we must get 802.1 involved so they can do
a bridge for this.

6. The packet handling components of the system, other than the file server
hosts, does not have a problem handling 1500 byte frames. It is the small 64
byte frames which kills them. To reduce load on the system, increase the
minimum frame size to 128 and you reduce the maximum packet forwarding rate
by 50%. In crease the minimum packet size to 640 and any system designed to
make forwarding decision for 1GbE will be able to handle 10GbE.

Henry Ngai

----- Original Message -----
From: Joe Gwinn <gwinn@xxxxxxxxxx>
To: Booth, Brad <bbooth@xxxxxxxxxx>
Cc: HSSG_reflector (E-mail) <stds-802-3-hssg@xxxxxxxx>
Sent: Monday, June 21, 1999 12:38 PM
Subject: Jumbo Frames in 10GbE?


>
> At 4:24 PM 99/6/17, Booth, Brad wrote:
> >Just a small point.  One of the objectives that passed with greater than
> >75% in Coeur d'Alene was to "preserve minimum and maximum FrameSize of
> >current 802.3 Std."
>
> I don't know that the issue is going to stay decided all that long, based
> on the recent article "Jumbo Frames gather support" (Jeff Caruso, Network
> World, 14 June 1999, page 6), which states that IETF has published a
> working document proposing that ethernet frames be made larger than the
> current 1,500-byte maximum, the basic rationale being to reduce the packet
> rate and thus load on packet-handling components of the system.  In short,
> this is a system issue, and cannot really be decided solely at the MAC
> level.
>
> If jumbo frames are to come, 10GbE would be a logical place to start.
>
> The issue will ultimately be decided by an IEEE Ballot Group, not a
> Plenary.  If the market is really going to bigger packets, as this article
> implies, it will be hard to resist.
>
> Joe Gwinn
>
>
>
> The above is in response to the following:
>
> >        >Issues 3 - Bit Error Rate
> >        >The assumption will be that this is 10-12. If someone wishes to
> >challenge
> >        >this they should bring a presentation to the next meeting
providing
> >        detailed reasoning why this needs to change.
> >
> >        It strikes me that the issue of larger maximum packet sizes will
likely
> >        come up, just as it did for GbE.  If 10GbE goes to 9 KB packets,
> >the design
> >        center BER would need to go to 10^-13 to maintain the same
theoretical
> >        packet loss rate.   I'm not sure how much effect this would have
in
> >        practice, as most gigabit links achieve much better than 10^-12,
> >if they
> >        work at all.  Anyway, these items are ripe for debate and
decision.
>
> **** end of message ****
>
>
>