[prev in list] [next in list] [prev in thread] [next in thread] 

List:       ipcop-user
Subject:    Re: [IPCop-user] IPCop with no hard drive (Eric Heflin)
From:       Melvin Backus <melvin () sleepydragon ! net>
Date:       2003-08-30 16:18:01
[Download RAW message or body]

Many, although apparently not all, of the Pentium Pros are able to run 
without any processor fan as well.  They have a HUGE heat sink and make a 
very good choice for unattended operations such as this.

At 02:03 PM 8/30/2003 +0100, Ged Haywood wrote:
>Hi all,
>
>On Sat, 30 Aug 2003 Chris M wrote:
>
> > > <snip>
> > > It would be far easier to just stick in an old drive.
> >
> > My objective is to remove / disable the IDE hard disk.  I expect it will be
> > the first component to fail under 24x7 operation.  Also, the hard disk
> > consumes lots of power, and is noisy.
>
>I speak from over thirty years of experience with electronic equipment. :(
>
>The components which will probably fail first are the fans.  These are
>by a very long way the least reliable parts of your computer.  The
>little fans on many of today's processors have the worst reliability
>record, again by a long way.  There is little difference as far as I
>can tell between 'sleeve bearing' fans and those which claim to have
>ball bearings.  I use the term 'claim to have' advisedly.  Next after
>processor fans and power supply fans, other parts of the power supply
>and anything else that can get hot or stressed are suspect too.
>Numerous power supplies are to be found in my electronics graveyard.
>
>You can turn off many hard drives from software to save power, and
>Linux supports this.  This may be an alternative for you.
>
>Somewhat surprisingly, if they are not mistreated then hard drives are
>much more reliable that one might expect.  I run dozens of hard drives
>continuously, and I haven't had one wear out for years.  My experience
>is that if they're going to fail it's usually quite soon after they've
>been installed.
>
>The last drive which failed for me (fell to pieces inside) was one I
>carried 160 miles in a tail pack on my Hyabusa (that's a motorcycle).
>It was one week old.  Perhaps it was to be expected, the same thing
>happens to my sunglasses if I put them in the tail pack.  At least the
>bike's not as bad as my friend's Harley-Davidson, he says the Harley
>dismantles his sunglasses while he's wearing them...
>
>One of the most important features to me when choosing a machine that
>may be going to run unattended (and perhaps not easily accessible) for
>a long time is convection cooling, especially of the processor.  In a
>PC this means you're limited to something like a 50MHz 486, but that's
>fine for a low-end Linux-based firewall.  If you want to be creative,
>use a length of four-inch drain-pipe to make a chimney to exhaust the
>power supply even if the PSU fan fails (or is removed:).  You'll need
>at least three metres of pipe, and obviously it needs to be largely
>vertical.  Avoid any sharp bends in the pipe.  You can make a large
>airtight box on the back of a PC to connect the pipe.  Make more holes
>in the case to let the air flow in.  If you site the holes with care
>and use a bit more ducting inside the case, then with a good heat-sink
>which is well-coupled to the CPU, you might even get away without a
>fan on a pentium-class processor.  A piece of four-inch drain-pipe can
>have a very large MTBF, so the main problem you will have then is
>dust, and especially airborne fibres which will eventually fill up
>your PC with dirty fluff.  Have you ever taken one to pieces after
>five years in a busy sales office?  Yuck.  Fetch the vacuum cleaner.
>
>If any of this sounds a bit Heath Robinson (yanks: Rube Goldberg) then
>consider that although I can afford the best equipment, I'll generally
>choose the simplest I can use.  I have several machines the have been
>running continuously, without being rebooted, for more than two years.
>Many of them have been running for more than five years, although they
>have been rebooted several times.  With some sadness, recently I took
>one out of service after thirteen years.  It's still working fine, but
>it's just too slow.  It has a 40Mbyte drive, the same one since new,
>which has run almost continuously all that time.  It's had at least
>two replacement power supplies, might be three.
>
>I'm not suggesting any of this might be right for you, but I think you
>might want to to consider probability, the laws of physics, and a few
>more options.
>
>73,
>Ged.
>
>
>
>-------------------------------------------------------
>This sf.net email is sponsored by:ThinkGeek
>Welcome to geek heaven.
>http://thinkgeek.com/sf
>_______________________________________________
>IPCop-user mailing list
>IPCop-user@lists.sourceforge.net
>https://lists.sourceforge.net/lists/listinfo/ipcop-user

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Whom computers would destroy, they must first drive mad.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Melvin Backus
Principal Wizard
Sleepy Dragon Enterprises
www.sleepydragon.net



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
IPCop-user mailing list
IPCop-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ipcop-user
[prev in list] [next in list] [prev in thread] [next in thread]