The IP over ATM Mailing List Archive by date

[Andrew Smith <asmith@Baynetworks.COM>]

> To: Andrew Smith <asmith@BayNetworks.COM>
> Cc: curtis@ans.net, ip-atm@matmos.hpl.hp.com
> Subject: Re: The purpose of IP over ATM and What size is a LIS?  
> Date: Thu, 11 May 1995 16:35:05 -0400
> From: Curtis Villamizar <curtis@ans.net>
> 
> In message <9505111845.AA08270@milliways>, Andrew Smith writes:
> > 
> > I don't think we should let this thread die until we have answered the
> > question posed by Grenville: the WG has never defined what it means by
> > what Curtis calls "large". The only concensus we seem to have so far
> > is that 2 <= "large" <= less-than-global. Can we please try to get
> > a little more focussed on narrowing this range? 
> > 
> > Andrew

> With all respect, I don't think continuing this thread is productive
> and I'll say why.  First let me summarize.  (Please correct me if I
> misquote anyone).
> 
> Grenville suggested that there were two dimensions of "large" when
> applied to discussion of "large LIS".  One was the number of hosts.
> The other was the geographic size, which then has implications on the
> number of switches in the LIS.  
> 
> I responded that 1) I was referring to "large" in terms of the number
> of hosts in discussion of protocol scaling and 2) that large in terms
> of the number of switches and delay between them is a traffic
> management issue which I don't think we want to rehash (please stop at
> http://engr.ans.net/nap-testing/ if before opening a discussion on
> this topic and maybe let do it off the list).

Grenville discussed (2) in his answer -I would just add that if you 
assume LISs are going to be kept local in all cases as opposed to split 
across large geographical distances then I think you are wrong, or else
you are seriously limiting the applicability of this protocol. Do people
really want to pay for 100 overhead VCCs for 100 hosts in the LIS across those 
long links, compared to 1 overhead VCC if they could deploy MCSs ??

I was looking for clarification on what you meant by (1). 

> I don't think it is useful to start arguing over whether the optimal
> LIS size is 50, or 100, or 200, or 1,000 hosts.  The point is that as
> the LIS grows, some things become problematic, an we've enumerated
> some.  If you try to pin a number of this, you get into issues such as
> are these 20 Crays or 20 Macintosh cumputers and what mix of protocols
> are going to be used.

The point is that if you want to build an LIS with today's ATM technology,
it *is* a critical issue to know if your system can deal with LIS sizes 
of 50, 100, 200 or 1000 hosts. You need to engineer your network very
differently if LISs cannot be built larger than 50 hosts - you might
actually want to contact your local router vendor if that were the case.
If LISs can be built to support 1000 hosts, you might only want to talk
to your favourite switch vendor and not bother sending all your sysops to
router school. Would you as a LAN sysop want to buy an ATM router port for every
few users (they're only $20000 or so each), in addition to that whole heap of
ATM switch hardware and software that you just bought?

[ Why am writing this? Profit margins on routers are nice and big - much
better than on ATM switches. I'll probably get shot at dawn tomorrow.... :-(]

Simple maths coupled with observation of the capabilities
of call processing systems would show you that a full mesh with more than 
a 100 or so endpoints split over several switches is going to be disastrous:
try routing 10000 Q.2931 calls at system startup or in failure modes
just to support the broadcast infrastructure. This will not provide 
customers with the sorts of service resilience they expect from data 
networks. If you tried peddling this sort of solution to LAN customers 
you would get laughed at - an reel of yellow co-ax can perform better than 
that!

You have to pin a number on the design goal this protocol has for the 
size of LISs due to today's state of the art - that is a part of the 
applicability statement. We know already that *you* are not interested in 
deploying LISs bigger than a few routers in size. I know that *I* am 
interested in deploying LISs with 100 hosts on them. Are there any other 
people out there listening with any views on this? Are you all building 
proprietary solutions? If so, I will shut up and go and implement my own 
MCS option and at least my customers will be able to use this technology.


> Curtis
> 
> ps- now can we end this thread...

Maybe soon, not yet, sorry.



Andrew


********************************************************************************
Andrew Smith					TEL:	+1 408 764 1574
Technology Synergy Unit				FAX:	+1 408 988 5525
Bay Networks, Inc.				E-m:	asmith@baynetworks.com
Santa Clara, CA
********************************************************************************