Date: Fri, 1 Nov 2002 10:29:30 -0500
From: Jeffrey Haas <jhaas@nexthop.com>
To: Qiu Jian <qiu@ns.6test.edu.cn>
Cc: "idr@merit.edu" <idr@merit.edu>
Subject: Re: Why MinRouteAdverInterval is recommended 30
Message-ID: <20021101102930.B13231@nexthop.com>
References: <20021101073553.4856E5DE1A@segue.merit.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.2.5i
In-Reply-To: <20021101073553.4856E5DE1A@segue.merit.edu>;
 from qiu@ns.6test.edu.cn on Fri, Nov 01, 2002 at 03:37:42PM +0800
Sender: owner-idr@merit.edu
Precedence: bulk

On Fri, Nov 01, 2002 at 03:37:42PM +0800, Qiu Jian wrote:
>     Is it right to say that the Maximal Propagation 
> Delay in Internet is almost less than 30 seconds, 

This is a matter of some debate.

> so MinRouteAdverInterval is set to 30 seconds to 
> guarantte the furthest route change can be caught 
> before sending the relevant update?

Theoretically, this means that if a route is still reachable (feasible),
then you will delay sending changes to the route.  This will
cause the entire routing system (in this case, possibly the Internet)
to have potential for blackholes and forwarding loops due to these
timers for some period of time.  The following paper goes into 
some details on this:

http://www.acm.org/sigcomm/sigcomm2000/conf/paper/sigcomm2000-5-2.pdf

-- 
Jeff Haas 
NextHop Technologies