🚀 go-pugleaf

keelers@igs.net (Stephen Keeler) wrote:

(snip)
> terminated (I'm assuming that the AHA-2940W is passively terminated 
> on-board -- please correct me if I'm wrong there!)

default is active termination enabled onboard.

> If better quality internal cable and/or active termination on an 
> internal cable is a suggestion, I'd also appreciate pointers to 
> companies (and even part names/numbers).  At this point, I'm a little 
> leary of spending even more money on hardware that 'might' fix the 
> problem.

Pat Davis of Division of Information Technology, UW-Madison advised
me to avoid the ribbon cables since he had problems with them in 
Novell servers he built.  He uses an IBM32G3922 which is twisted pair
ribbon except where the connectors go on.  It isn't cheap but it is
working for me, on Seagate 15150W's.

I tested it with a round-robin copy of a multimegabyte file.  After
about 4GB worth of reading and 4GB of writing, there were no errors
found when comparing the final copy to the original.
file O ->a -> b -> c -> a -> b -> c etc for about 800 generations, 
then compare to O)

Another possibility is Granite cables and terminators.

Good luck.

Ken