Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 12 Sep 1996 15:02:21 -0700 (MST)
From:      Terry Lambert <terry@lambert.org>
To:        fenner@parc.xerox.com (Bill Fenner)
Cc:        karl@mcs.net, terry@lambert.org, avalon@coombs.anu.edu.au, freebsd-hackers@freebsd.org, koshy@india.hp.com
Subject:   Re: SYN Resisting (fwd)
Message-ID:  <199609122202.PAA07685@phaeton.artisoft.com>
In-Reply-To: <96Sep12.141252pdt.177595@crevenia.parc.xerox.com> from "Bill Fenner" at Sep 12, 96 02:12:50 pm

next in thread | previous in thread | raw e-mail | index | archive | help
> >Like you are probably ignoring source quench.
> 
> Source quenches are deprecated by RFC1812.  RFC1122 is still the host
> requirements RFC so technically hosts must still pay attention to them,
> but routers are not supposed to send them.
> 
> 4.3.3.3 Source Quench
>    
>    A router SHOULD NOT originate ICMP Source Quench messages.
> ...
>    DISCUSSION 
>       Research seems to suggest that Source Quench consumes network
>       bandwidth but is an ineffective (and unfair) antidote to
>       congestion.  See, for example, [INTERNET:9] and [INTERNET:10].
>       Section [5.3.6] discusses the current thinking on how routers  
>       ought to deal with overload and network congestion.

I don't know the wire protocol in use, but I would suspect that if it's
LD cable transport, that it's ATM without a committed rate.  This would
be, by definition, unreliable small datagram delivery (aka "leaky bucket").
ATM manages loss conditions not by committing rates, but by issueing
source quench base on dropped packet counters.  Yes, I know, ATM is
annoying, but it is what the phone companies use for large links, and
we are stuck with it.

Other than that, I was a little peeved at blaming the US with the blanket
statement that the loss was on the US end of things.  Ignoring perfectly
valid source quench requests (from *non*-ICMP ATM routers) is only one
of the possibilites that could be considered before calling everyone
managing NSP in the US incompetent.

Frame Relay has the same source quench issues as ATM (or IPV6, for that
matter), so it;s not clear where the blame lies.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199609122202.PAA07685>