[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: First Beowulf cluster. Some pointers?

> The disk less setup I did a couple of years ago only had a special boot floppy,
> which had a DHCP/BOOTP kernel dd'd on it ('rdev /dev/nfs' or something...)
> which mounted / from a NFS server... It was slow, but worked find.

We had problems with /usr/lib, and a few other directories as well. I
don't recall offhand which ones, since it's been a while.

> Clustering 10 machines (the reasonable amount of SPARC stations I could
> get working I think) would equal a machine with 500 bogomips...

Ok, first off, you won't get that. You've got overhead on each machine,
plus the overhead of the message passing to contend with.

> THAT is what I would like. I put the master node as the front, and have the
> slaves behind that (somehow) and you login to a machine that is just a
> simple 486/33, but with the processing power of, what? A PII/200 (or whatever
> it would result in)?

Second off, there's a massive difference in operational capability from an
SMP/parallel box to a single proc box. Different stuff runs well on each.
> I might have misunderstood exactly what a cluster (or Beowulf cluster perhaps),
> but a couple of years ago a friend (same friend as in previous mails) had a
> HP 300/something... There where three nodes, one disk. I don't think I ever
> saw this running, but I was told "it's one machine, with the load spread to
> three CPU's"...

Not true. It isn't one machine, it's several machines operating in tandem.
Think of it like that, rather than one large box. It's operational uses
are more specialized than a singe machine.


Reply to: