[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Reopening jumbo frames in IS-IS

:^)

next time you're running on a line in which a 53 byte frame is necessary to avoid jitter issues, please let me know. I'll be interested to know that dinosaurs still roam the earth. Hadn't actually noticed any since my childhood...

:^)

The reference to timing has to do with T-1 and E-1 lines and their slower cousins. It's not much of an issue on a 10 MBPS channel, and is only an issue on 2 MBPS links when queues back up (12000 bits at 1.544 MBPS is a tad shy of 8 ms). But when one gets on the 64 KBPS lines that are still so common in many parts of the world, a 12,000 bit packet is 187 ms, and we jump through hoops to make other applications work well. Now repeat the math using 72,000 bit frames.

And then we could discuss the price of a SAR at 10 GBPS etc. The 53 byte frames become a little pointless at anything resembling optical rates.

On Jul 15, 2005, at 3:09 PM, Manfredi, Albert E wrote:

-----Original Message-----
From: Bob Hinden [mailto:bob.hinden at nokia.com]

p.s. I also find it ironic that in the other side of the
house there are
folks who think 53 byte packets are ideal :-)


Hmmm. Didn't someone recently mention that going to biiger MTUs than
1500 would cause potential jitter problems, as longer frames got in
queues with shorter ones?

Quoting:

"I can't speak for IEEE, but the reasons usually brought up include
implementation costs in terms of buffer depths, and mutual jitter
between competing traffic streams. If you have a session that sends a
packet every 20 ms and depends on that being mostly maintained,
having another session send packets that are 30 ms long and can get
several into the queue ahead of you can be a real pain."

Seems like this is where the 48-byte crowd was coming from too, although
they wanted to be able to go up and down in link speed without having to
change the size of the "frame."

;)

Bert


_______________________________________________
routing-discussion mailing list
routing-discussion at ietf.org
https://www1.ietf.org/mailman/listinfo/routing-discussion