Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site ames-lm.UUCP
Path: utzoo!watmath!clyde!akgua!sdcsvax!sdcrdcf!hplabs!ames-lm!eugene
From: eugene@ames-lm.UUCP (Eugene Miya)
Newsgroups: net.lan
Subject: Re: VAX to SUN via Ethernet
Message-ID: <244@ames-lm.UUCP>
Date: Fri, 11-May-84 12:11:48 EDT
Article-I.D.: ames-lm.244
Posted: Fri May 11 12:11:48 1984
Date-Received: Sun, 13-May-84 08:27:10 EDT
Organization: NASA-Ames Research Center, Mtn. View, CA
Lines: 15

We have a SUN (1.5) connected to three other VAXen via Ethernet
(The VAXen are also connected via HYPERchannel which we are also
trying to connect the SUN.)  I did some preliminary test work on
throughput using 4.1c/4.2 at the turn of the year.  I am only now able to
get a TR out on performance testing, too late for Usenix, Maybe in
January.  Anyway, the performance of Ethernet is better than Hyperchannel
on 4.2BSD.  We are certain yhe problem has to do with the size
and complexity of the HYPERchannel/Hy driver. Ethernet/TCP also has
some somewhat unexpected performance peaks (Best buffer size is 1024 bytes
for a stand along thru put of ~90 KB/sec.

more later.

--eugene miya
  NASA Ames Res. Ctr