Mailing-List: contact tomcat-user-help@jakarta.apache.org; run by ezmlm
From: "Alexandros Kotsiras" <kotsiras@mediaondemand.com>
To: <tomcat-user@jakarta.apache.org>
Subject: RE: Can i start/stop Tomcat from a remote computer ? ? 
Date: Wed, 26 Jul 2000 20:26:05 -0400
Message-ID: <BCAD6DE3E75AD311827D00C04F609B9C06F882@MOD_EXCHANGE>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
Importance: Normal
In-Reply-To: <BCAD6DE3E75AD311827D00C04F609B9C7EDD89@MOD_EXCHANGE>

Well , i am kind of new to the Computer business at least professionally  so
i am not really aware of the
old days but i enjoy the comments, at least i learn something.
Based on my NOT really in-depth JAVA knowledge i though that 4MB is "light"
for JAVA mainly for 2 reasons :

1. I remember once reading that whatever JAVA app you code your memory usage
starts at 4MB or higher because this (4MB) is what the JVM needs by itself.
Is that correct ? ?

2. Whenever i start Tomcat on WinNT, exactly as it comes from
the distribution without adding contexts and additional load-on-startup
servlets i see the
java.exe process on WinNT Task Manager going to +8MB.

So i thought why should i use a second Admin Tomcat Instance (+8MB extra)
like most of the Commercial products do (JRun)
in order to start/stop the Application instance ? ?
And i was kind of happy to see my  server-socket process which does what i
need going "only" up to 4MB-4.5MB when it starts.

Anyway ..i would really like to know whether my first assumtion about the
4MB needs of the JVM  is correct..

Alex.

-----Original Message-----
From: John Summerfield [mailto:summer@OS2.ami.com.au]
Sent: Wednesday, July 26, 2000 7:17 PM
To: tomcat-user@jakarta.apache.org
Subject: Re: Can i start/stop Tomcat from a remote computer ? ?


> Alexandros Kotsiras wrote:
>
> > It has a lightweight server-side part (approx 4MB)
>
> I always smile when I read things like this ... developers nowdays are so
> SPOILED :-) :-) :-)
>
> The first "mainframe" that I programmed on professionally had 32 KILObytes
of
> main memory, and we couldn't ever write a complex enough program (in RPG
II)
> to
> use it all effectively ... :-)


I can't resist the urge. Please forgive me;-)


The first computer O got at (I was a beginner then) was a CDC 3200, 16K
24-bit
words.

By first program, in Fortran, overwrote the operating system, SCOPE.

Social Security (Australia) implemented Medibank in the mid 70s on an IBM
mainframe with
6 Mb RAM
OS/VS2 Rel 1 (16 Mbyte aggregate [single] address space)
A brace of disk drives (3330), 100 Mbytes each.
6 disk drives (2319), 29 Mbytes each.
A dozen or so 3420 (tape) drives.
3 3211 priters (on their own floor)

Each Mbyte of RAM was in its own box; these boxes were about two metres
high.

The computer, a S/370 model 168, could sustain 1.5 Mbytes/second on a block
multiplexor channel (approximately equivalent to a SCSI controller), and 3
Mbytes/second all up.

There were actually two of them; the second had 2 Mbytes and didn't do much.

Subsequently they were configured as a single multiprocessor system under
OS/VS2 3.0 (MVS; each job had its own 16 Mbytes address space).

One of the really cool features was its dynamic reconfigurabiity. One
evening
I dropped in and asked if I could have one to do wome work (using OS/VS1).
We
configured out a Mbyte of RAM and a CPU and a disk drive and a console and
threw a switch or two and I had myself a computer, \. Meanwhile everything
else kept right on running; when it was booted, it was running on an MP
system, now it was on a UP (uniprocessor). When I finished, it was all put
together and MVS was running on an MP system again.