Re: [tlug] SATA software RAID or SAS hardware RAID?

[Edward Middleton <edward@example.com>]

Stephen J. Turnbull wrote:
> Also, while I'm no expert, getting multiple anything to perform better
> than a single piece of hardware seems to be non-trivial.

This depends on the problem.  Things that can use a standard interface
like BLAS or LINPACK, for all the heavy work, can be trivially
converted, and if the problem is not inherently serial in nature you can
get orders of magnitude increases in performance.  At university I got
orders of magnitude increase in performance doing numerical modeling by
moving from the f77 to the SUN tool chain which was heavily
multi-threaded.  Because I was using a standard interface the conversion
was trivial.

> Eg, I was
> surprised but not shocked to discover that a benchmark program that in
> sequential mode takes 1.2 seconds for 1000 iterations takes over a
> second *per iteration* in parallel mode[1] on a 2 x Dual Opteron
> system.
>   

If you are only running on a single machine why are you using PVM?

>   My suspicion is that the bottleneck is master-slave messaging
> from socket to the other (I know the bottleneck is intermittent, and
> it is in the messaging, but I haven't confirmed that it matters which
> pair of CPUs are involved).  The breakage is pretty spectacular, even
> though the parallel monitor is very-well-respected software, and the
> mobo (Tyan 2895) and CPUs are excellent.
>   

Perhaps it is not optimized for parallel processing on a single machine.

> Footnotes: 
> [1]  Using the PVM package, which probably communicates via TCP/IP
> over Unix-domain sockets.  MPIv2, several implementations of which
> have good shared memory support, seems to be the way to go for real
> work, but PVM comes with "batteries included" for demos, eg, a nice
> XPVM visual monitor, and extending to heterogenous networked VMs is
> trivial.
>   

Edward