[3599] in linux-net channel archive
Re: sna and bridging
daemon@ATHENA.MIT.EDU (Matti E Aarnio)
Tue Jul 9 04:58:12 1996
From: Matti E Aarnio <mea@mea.cc.utu.fi>
To: niemi@wauug.erols.com (David C Niemi)
Date: Tue, 9 Jul 1996 11:18:43 +0300 (EET DST)
Cc: devel@vrml.k12.la.us, linux-net@vger.rutgers.edu, chessman@wauug.erols.com
In-Reply-To: <Pine.LNX.3.94.960704002005.5245C-100000@wauug.erols.com> from David C Niemi at "Jul 4, 96 00:42:31 am"
David C Niemi speaks about SNA and Linux
...
> However, if this is on your firewall, the other SNA box is probably not
> sitting just on the other side of the Linux box, and SNA will not route
> too well over the Internet (!). So you probably really need SNA
> tunnelling over IP (aka encapsulation), and you'll need a gateway to each
> SNA host (which could in theory be the SNA hosts themselves or their
> minions who attach them to the LAN, if they are smart enough). It is even
> possible to convince Big Iron to speak TCP/IP, though admittedly with a
> thick accent.
Well, not so bad an accent, although it is already many
years since I spoke that accent daily (with a Big Iron)..
Last I heard, the price of the TCP/IP for Big Iron is HUGE!
Still, the lattest IBM machines are also marketed to be
serious competition at transaction market -- where the
system IO performance counts.
> We do this where I work, and FTP to the mainframe is almost as fast as
> between two Linux boxes with $8 ethernet cards, which is to say it is at
> least an order of magnitude faster than proprietary transfer schemes
> over LU 6.2 bisync connections (and also far more reliable).
Well, the BISYNC is awfull connection method -- half-duplex,
and all that. Still, the BITNET is made by using it, and for
a while it was the major network in the world (Anybody remembers?
Everybody are such a young people that they have seen only the
Internet ? -- in fond memory of our Big Iron: FINTUVM)
(FYI: There are a few UNIX machines at the BITNET with my
FUNETNJE software, at least one of them is Linux machine..)
> I have not heard much action on SNA in the Linux community, probably
> because there is major culture shock between blue-suited big-iron fans and
> mellowed-out sandal-and-tshirt-wearing Linuxers. It would probably not be
> all that difficult technically, though, if you want to take it on ;^)
Actually it is technically difficult. I have heard several
accounts that SNA is not specified fully at its documentation,
and that each new VTAM release that IBM puts out has its own
"features" ( that is, deviations from the specs -- call them
bugs, if you like ) which must be catered for, and only way
to learn about them is by using protocol analyzers, and to test...
> David
> Niemi@wauug.erols.com 703-810-5538 Reston, Virginia, USA
> ------ Money talks, but it is wrong half of the time. -----
/Matti Aarnio <mea@utu.fi> <mea@nic.funet.fi>
<mea@finfiles.bitnet> :)
