tracym@nsd.3com.com Thu, 13 January 1994 23:20 UTC

Errors-To: catnip-request@world.std.com
Reply-To: catnip@world.std.com
Precedence: bulk
Message-Id: <199401132204.AA17395@remmel.NSD.3Com.COM>
To: catnip@world.std.com
In-Reply-To: Your message of "Wed, 12 Jan 94 17:13:22 EST." <199401122213.RAA17983@titan.sprintlink.net>
Date: Thu, 13 Jan 1994 14:04:34 -0800
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: tracym@nsd.3com.com

> >It also seems that 8 bits is really sufficient.
> 
> Not at 1/10 seconds!  The problem is that there are links which
> can delay a packet for up to a minute while establishing a connection.
> We'll see more and more of them with the proliferation of dial-up IP
> and the real cheap 19.2Kbps modems.
> 
> I would propose "logarithmic" time -> TTL translation; i.e.
> 
> 	1/16 sec delay equals 1 hop
> 	1/8 sec == 2 hops
> 	1/4 sec == 3 hops
> 
> etc.  I'm not sure what the implications of that scheme are.

Aside from possibly confusing traceroute, this looks ok.  The most
important point is that the field ALWAYS gets updated(decremented or
otherwise) by at least "1".

[A classic SPF routing update distribution bug is to modify
link-state-advertisement ages based on internal clock ticks without
ensuring that the age of a new-to-this-router LSA gets updated at
least once before being sent to other neighbors.  This has formed the
heart of some live-forever scenarios.]

Tracy

Vadim Antonov
tracym
Re: (it must be lost) Matt Crawford