From owner-freebsd-questions  Tue Sep 17 09:04:26 1996
Return-Path: owner-questions
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id JAA19277
          for questions-outgoing; Tue, 17 Sep 1996 09:04:26 -0700 (PDT)
Received: from diablo.ppp.de (diablo.ppp.de [193.141.101.34])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id JAA19266
          for <questions@FreeBSD.org>; Tue, 17 Sep 1996 09:04:20 -0700 (PDT)
Received: from allegro.lemis.de by diablo.ppp.de with smtp
	(Smail3.1.28.1 #1) id m0v32dC-000QlNC; Tue, 17 Sep 96 18:04 MET DST
From: grog@lemis.de (Greg Lehey)
Organisation: LEMIS, Schellnhausen 2, 36325 Feldatal, Germany
Phone: +49-6637-919123
Fax:   +49-6637-919122
Received: (grog@localhost) by allegro.lemis.de (8.6.9/8.6.9) 
	id RAA09223; Tue, 17 Sep 1996 17:57:33 +0200
Message-Id: <199609171557.RAA09223@allegro.lemis.de>
Subject: Re: Slow Etherlink
To: black@gage.com (Ben Black)
Date: Tue, 17 Sep 1996 17:57:32 +0200 (MET DST)
Cc: questions@FreeBSD.org (FreeBSD Questions)
In-Reply-To: <9609171541.AA07852@squid.gage.com> from "Ben Black" at Sep 17, 96 10:41:46 am
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
Sender: owner-questions@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

Ben Black writes:
>
>> The real surprise is Solaris 2.5.  The SS2 only has 16 MB of memory,
>> but all it was doing was receiving the file, so you'd think it could
>> handle things better than that.  Does anybody have any ideas?
>>
>
> 16MB is not nearly enough for a solaris 2.5 machine.

To be fair, that should depend on what you're doing.  But yes, it's
not normally enough.  Now, on the other hand, I'm compiling gcc on it,
normally an operation which needs a lot of memory, and it's not doing
too badly (except that it rebooted on me without saying why).

> it's a severely bloated bit of software.  things might get radically
> better if you went to 32MB on it.

That would assume that it's paging severely with 16 MB.  I don't think
that's the case, and it would be really bad if it were.

> also, are your benchmarks in bits or bytes per second?

Bytes, not bits.  It couldn't be that bad :-)  The high values (1030
kB/s) represent a little over 80% of the theoretical bandwidth.

Greg