[ previous ] [ next ] [ threads ]
 From:  Joey Morin <jmorin at icomm dot ca>
 To:  Dinesh Nair <dinesh at alphaque dot com>
 Cc:  Eric Shorkey <eshorkey at commonpointservices dot com>, Adam Nellemann <adam at nellemann dot nu>, <m0n0wall at lists dot m0n0 dot ch>
 Subject:  Re: [m0n0wall] Wondershaper
 Date:  Wed, 9 Jun 2004 15:11:23 -0400 (EDT)
This one time, at band camp, Dinesh Nair said:

> >    this is because many connection types (like HTTP) employ a scheme
> >    whereby acknowledgement packets (ACK) are returned for each packet
> >    or group of packets received.
> >
> >    so, an HTTP server won't send you the next packet in your huge
> >    download until it has received an ACK packet for it's last burst.
> err, chaps, that is part of the tcp protocol. all tcp connections follow
> that behaviour, be it http, ssh, smtp, pop and stuff. udp is a
> connectionless protocol and doesnt. udp examples include dns, RADIUS,
> and some p2p networks.

i never implied otherwise.  there are other protocols besided TCP and UDP,
i'm not familiar with all of them.  i expect some are ACKing, and others
are ACKless.  HTTP was just an example of the former.

> >    upload, or bittorent process) leaves little room for small packets
> >    generated by keystrokes in an SSH session, so you see some latency.
> the soon to be released magic shaper automagically creates a priority
> queue for small packets (length < 40bytes payload) to address
> interactive usage such as ssh, telnet et al.

will magic shaper and the existing traffic shaper co-exist in future
releases of m0n0?  i hope so.  i can see it would be nice to have things
happen magically, but i also like to get under the hood a bit...