Beowulf digest, Vol 1 #542 - 2 msgs
tlioner at 263.net
Fri Aug 24 05:02:51 PDT 2001
Send Beowulf mailing list submissions to
beowulf at beowulf.org
To subscribe or unsubscribe via the web, visit
or, via email, send a message with subject or body 'help' to
beowulf-request at beowulf.org
You can reach the person managing the list at
beowulf-admin at beowulf.org
When replying, please edit your Subject line so it is more specific than
"Re: Contents of Beowulf digest..."
1. Re:Channel Bonding (=?iso-8859-1?Q?Jakob_=D8stergaard?=)
2. RE:Network RAM for Beowulf (James Alton)
Date: Thu, 16 Aug 2001 20:55:49 +0200
From: =?iso-8859-1?Q?Jakob_=D8stergaard?= <jakob at unthought.net>
To: Jared Hodge <jared_hodge at iat.utexas.edu>
Cc: Rocky McGaugh <rmcgaugh at atipa.com>, beowulf at beowulf.org
Subject: Re: Channel Bonding
Jared Hodge <jared_hodge at iat.utexas.edu>,
Rocky McGaugh <rmcgaugh at atipa.com>, beowulf at beowulf.org
On Thu, Aug 16, 2001 at 09:19:19AM -0500, Jared Hodge wrote:
> I thought the MAC was determined by the NIC (which would mean it would
> vary from NIC to NIC even if in the same machine). I may be wrong.
It is. But you can overwrite it (both non-permanently as the bonding code
does, or permanently on many NICs using Donald Becker's tools).
MAC collissions can be catastrophical for normal networks, so usually you don't
want to fiddle with them. Since the NICs come with pre-configured unique MACs
there's no reason for most drivers to even think about changing it. The channel
bonding case is unusual.
: jakob at unthought.net : And I see the elder races, :
:.........................: putrid forms of man :
: Jakob Østergaard : See him rise and claim the earth, :
: OZ9ABN : his downfall is at hand. :
From: "James Alton" <jalton at olsh.cx>
To: <beowulf at beowulf.org>
Subject: RE: Network RAM for Beowulf
Date: Thu, 16 Aug 2001 12:36:21 -0700
Would it be possible to bypass the entire, network a pile of PCs approach?
Instead, could you build a circuit board where you plug in a hoard of 512MB
PC133 sticks (oh so $30 cheap now), run the board at 133 or faster,
interface through the PCI slot or something faster? It would run almost
infinitely faster than going through an Ethernet port. It's too bad I don't
know enough about electronics to build such a beast myself. I only know
that's you'd have to run the board at memory speed, have some memory
controller, perhaps a board controller to interface to the gigantic memory
space, and some support hardware (capacitors and resistors here and there).
If you made it interface though PCI, I don't think there would be a way to
map it directly into the system as I remember a 4GB limit of 32-bit
machines. Perhaps you could issue command to it instead and have the board
use a 64-bit system? In any case, you could use some type of hacked (because
of the 64-bitness of it) MD (malloc disk) to make a super fast file system.
All in all, I think something similar to this was approached using the IDE
interface, but IMHO, such a thing is much too slow for memory bandwidth.
(newer versions of PCI might be fast enough to get reasonable speed...)
jalton at olsh.cx
From: beowulf-admin at beowulf.org [mailto:beowulf-admin at beowulf.org]On
Behalf Of Amber Palekar
Sent: Thursday, August 16, 2001 2:35 AM
To: beowulf at beowulf.org
Subject: Network RAM for Beowulf
we as a group of four students are *also* thinking
of implementing Network RAM for a beowulf cluster
(assuming 100Mbps Ethernet ) whereby each node in the
cluster will donate some part of their RAM to be used
by all other nodes.so we will basically be mapping
this shared RAM to the address space of the current
node.One of the uses that we're thinking of is for
Journaling(as in file systems ).We'll be maintaining
the journals on the Network RAM instead of writing
them to the local disks.As we are completely new to
this , it is very difficult for us to determine the
statistics like :- the overhead in writing to Network
RAM . Any info or pointers to these stats would be
highly appreciated .
Do You Yahoo!?
Make international calls for as low as $.04/minute with Yahoo! Messenger
Beowulf mailing list, Beowulf at beowulf.org
To change your subscription (digest mode or unsubscribe) visit
Beowulf mailing list
Beowulf at beowulf.org
End of Beowulf Digest
More information about the Beowulf