Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 10 Sep 2000 16:09:40 -0400
From:      Laurence Berland <stuyman@confusion.net>
To:        sthaug@nethelp.no
Cc:        mike@mikesweb.com, wizard@sybaweb.co.za, freebsd-isp@FreeBSD.ORG
Subject:   Re: NIC settings
Message-ID:  <39BBEA84.5F306051@confusion.net>
References:  <39BBE2A6.76BA1145@confusion.net> <15790.968614746@verdi.nethelp.no>

next in thread | previous in thread | raw e-mail | index | archive | help
Let's see if this is the right way to describe the process 
1.  packet comes across t1 bound for local ethernet.
2.  the router examines the packet, determines it should go out over
ethernet.
3.  the router uses csma/cd and gets one of the followng
	a) line in use, so wait for it free
	b) line free, transmits frame successfully
	c) line free, transmits frame, collision occurs so back off a random
time and retry.

C is the one we're interested in.  I don't see how it would increase the
utilization of the T1 at all (as the author to whom I originally replied
alluded)

sthaug@nethelp.no wrote:
> 
> > Should not the router on the T1 be performing some sort of cacheing so
> > that packets don't get retransmitted across the T1 every time a
> > collision occurs?
> 
> Um, packets *don't* get retransmitted across the T1 every time a
> collision occurs. In fact packets don't even get *to* the router's
> Ethernet port until the collision is resolved.
> 
> Steinar Haug, Nethelp consulting, sthaug@nethelp.no


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-isp" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?39BBEA84.5F306051>