Return-Path: owner-postman
Received: from localhost.Berkeley.EDU (localhost.Berkeley.EDU [127.0.0.1]) by nobozo.CS.Berkeley.EDU (8.6.4/8.6.3) with SMTP id KAA19977 for postgres-redist; Thu, 14 Jul 1994 10:35:57 -0700
Resent-From: POSTGRES mailing list <postman@postgres.Berkeley.EDU>
Resent-Message-Id: <199407141735.KAA19977@nobozo.CS.Berkeley.EDU>
X-Authentication-Warning: nobozo.CS.Berkeley.EDU: Host localhost.Berkeley.EDU didn't use HELO protocol
Sender: owner-postman@postgres.Berkeley.EDU
X-Return-Path: owner-postman
Received: from RI.CMU.EDU (RI.CMU.EDU [128.2.222.162]) by nobozo.CS.Berkeley.EDU (8.6.4/8.6.3) with SMTP id KAA19967 for <postgres@postgres.berkeley.edu>; Thu, 14 Jul 1994 10:35:56 -0700
Received: from ARETE.EDRC.CMU.EDU by RI.CMU.EDU id aa23916;
          14 Jul 94 13:33:46 EDT
Received: by arete.edrc.cmu.edu; id AA14648; Thu, 14 Jul 1994 13:37:52 -0400
Message-Id: <9407141737.AA14648@arete.edrc.cmu.edu>
From: Sean Levy <snl@postgres.Berkeley.EDU>
Date: Thu, 14 Jul 94 13:37:51 EDT
To: Bodo.Parady@eng.SUN.COM (Bodo Parady - SMCC Architecture Performance	Group)
Cc: postgres@postgres.Berkeley.EDU
In-Reply-To: Bodo.Parady@Eng.Sun.COM's message of Mon, 20 Jun 1994 12:08:26 +0800
Subject: Re: Message passing version of Postgres
Reply-To: snl+@CMU.EDU
References: <+cmu.andrew.internet.postgres+Ai1SrQC00UfA00vsV:@andrew.cmu.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Resent-To: postgres-redist@postgres.Berkeley.EDU
Resent-Date: Thu, 14 Jul 94 10:35:57 -0700
Resent-XMts: smtp

We have looked at modifying postgres to use ISIS as its communications
subsystem instead of sockets.  The diffs would be rather complex, and
we have some pretty severe time constraints, so we backed off and are
now pursuing a strategy of leaving postgres itself alone, but some
server programs which are accessed via ISIS and which themselves talk
to postgres via the normal libpq mechanisms.  Our implementation is
not quite up yet, but it looks like it is a workable solution although
not as scalable as it would've been if ISIS had been in there at a
lowlevel.  As PVM doesn't really address scalability, though, this
probably wouldn't be an issue for you.  In principle, the same
approach could be taken with PVM for a "cheap" first shot.  I'm not
familiar with MPI.

Good luck,
				-- Sean

==============================================================================
   To add/remove yourself to/from the POSTGRES mailing list: send mail with 
   the subject line ADD or DEL to "postgres-request@postgres.Berkeley.EDU"

   If this fails, send mail to "post_questions@postgres.Berkeley.EDU" and
   a human will deal with it.  DO NOT post to the "postgres" mailing list.
==============================================================================