Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!uakari.primate.wisc.edu!ginosko!gem.mps.ohio-state.edu!tut.cis.ohio-state.edu!ucbvax!msa3b.UUCP!kevin
From: kevin@msa3b.UUCP ("Kevin P. Kleinfelter")
Newsgroups: comp.protocols.ibm
Subject: Re: TCP/IP over SNA
Message-ID: <8909261253.AA07662@jade.berkeley.edu>
Date: 25 Sep 89 23:47:35 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Reply-To: "Kevin P. Kleinfelter"
              
Organization: The Internet
Lines: 43

robert@peregrine.UUCP (Robert Young) writes:

>In article <1079@ks.UUCP> drake@ibmarc.UUCP (Sam Drake) writes:
>>This isn't a commercial, but ...
>>
>>IBM's VM and MVS TCP/IP products can transport TCP/IP over SNA networks,
>>and there's a PRPQ available from IBM to allow the RT to do the same.
>>
>>Sam Drake / IBM Almaden Research Center

[Stuff about connections suggested by IBM omitted.]

>So far it looks like the IBM people we deal with are not completely sure
>about any clear solutions.
>Any clues would really be appreciated.

O.K.  I have spent quit a bit of time hashing this out with many different
people, and this is how it looks to me:

1. You CAN send TCP/IP over SNA -- in SOME cases.
   To attach SNA to TCP/IP (using AVAILABLE hardware/software) you
   MUST have a host (i.e. a 370) and a host adapter (I think it is an
   8232).  To route TCP/IP over SNA would look as follows:

workstation---8232--370====(SNA NETW)===370--8232---workstation

There are plans involving RT's and SDLC links, but they do not seem
(to me) to really be routing TCP/IP over an SNA network.

2. The "normal" IBM communications hardware (3705, 3174, etc) do
   not support TCP/IP connections.  You have to attach to a host
   at each end of your link.  Future versions of the 3705 and 3174
   will probably support TCP/IP, according to an IBM employee
   I met at an IBM class, but that is "unofficial."

3.  We don't like the above.  We are looking at multiplexing TCP/IP and
    SNA over our existing lines, or using an X.25 service, or
    using other alternatives to running TCP/IP over SNA.  If you find
    a better solution, PLEASE POST!  (Other people seem to be interested!)

--
Kevin Kleinfelter @ Management Science America, Inc (404) 239-2347
gatech!nanovx!msa3b!kevin