Re: anybody tried this?

Jason Williams (streak@ccwf.cc.utexas.edu)
Thu, 25 Jun 1998 10:45:53 -0500 (CDT)


On Wed, 24 Jun 1998, Danny Morris wrote:
> Far be it for me to say anything to contradict Jason, but from my
> understanding, it has alot to do with the max trans. rate that your IP
> handles.

Feel free to contradict me....I'm not an expert in TCP/IP theory. :)

> MaxMTU lets you change and test different packet (it is made for
> ppl that dont understand all of this stuff, like me :-) sizes and shows
> weather or not they where fragmented when you ping a server.

Where can one obtain MaxMTU?

> The standard is MTU=576, here anyway, when I tried higher it would
> fragment the packets

I just always thought the standards was 1500 since, I believe, that's what
most routers are set to. Limiting it to 576 at the end points doesn't
really help anyone so I'm not sure why it's done.

Ideally what you want is a program that determines the MTU settings for
each point along the connection (ala traceroute style).

> correct me here Jason, could just slow down your transfer rate.

I'm not sure what it would do since I haven't seen it happen...From others
that I've talked to, they said their wasn't much difference in CU when
messing with the MTU setting. I'd think it'd have a lot more to do with
the quality settings. If your MTU is 576, you'd want your frame size to
be a multiple of that I'd think. If you're sending with a frame size of
1300 or something, the end point has to assemble multiple fragmented
packets per frame. The reflector can adjust the sending rate and
receiving rate of each participant, but it can't adjust the quality of the
participants. It ultimately shows up as high packet loss on CU.

--
streak@ccwf.cc.utexas.edu    * Jason Williams -- Austin, Tx.  |     |
streak@mail.utexas.edu       * University of Texas at Austin  | ___ |
streak@cs.utexas.edu         * BS Computer Science             \_|_/
*************** http://ccwf.cc.utexas.edu/~streak/ **************|