Cluster Upgrade Strategy

[Dean Johnson]

On Thu, 2002-10-17 at 16:57, Ed Hall wrote:
> Folks,
> 
> I have some fairly basic questions about upgrading/expanding a existing
> (less than 1 year old) Linux cluster with additional nodes. The cluster
> is used in a University environment for both serial and parallel processing.
> 
> Is it better to buy older cpus that are the same as on the existing nodes
> to keep the cluster homogenous so that its easier to administer and use?
> Or does it make more sense to buy newer (and possibly fewer) cpus with
> greater compute capacity at the expense of having a uniform set of nodes
> in the cluster? If the only alternative is to buy newer nodes, should they
> be used as a separate cluster rather than combining them with an existing
> cluster of older nodes?
> 

It all depends on what you are doing and how your algorithms are
structured. If you have a simplistic work unit model with a round robin
work unit scheduling strategy, then better nodes are probably not
terribly useful. If you can effectively deal with work units taking a
range of completion times, then get the fastest nodes you can or the
best bang for the buck.

	-Dean