Bonded head nodes

[Donald Becker]

On Fri, 8 Nov 2002, Ivan Pulleyn wrote:
> On Fri, 8 Nov 2002, Donald Becker wrote:
> 
> > Every current motherboard you might use in a cluster has native PXE
> > boot.  It's not the protocol you would choose, but it's a ubiquitous
> > standard.  It's done.  It's workable.  That means it won't change for
> > about a decade.
> 
> I agree that stage 1 will remain PXE w/TFTP for a long time because
> it's burned into the ROMs. However, the bulk of the data transfer
> happens afterwards (due to the transfer size constraints imposed by
> the ROMs). There's no reason we can't work as a group to improve stage
> 2 performance. This merely requires getting new code accepted into the
> various Linux projects that implement stage 2 clients.

We did that work well over 2 years ago, and published the source.  There
is no "work as a group" to be done unless you mean "crib the code so that
I can get credit for the idea".

-- 
Donald Becker				becker at scyld.com
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Scyld Beowulf cluster system
Annapolis MD 21403			410-990-9993