SATA vs. SCSI drive performance?

[CA789823]

Are there any tests that compare the HP NAS 1500 using (4) SATA drive vs. the HP NAS 2000 using (4) SCSI Drives?

If the performance is about the same, I would rather opt for the NAS 1500 (4 drive) 250GB SATA setup as opposed to the NAS 2000 (4 drive) 146GB SCSI solution.

Also, does adding RAM to the standard NAS config really buy you much in the way of overall performance?

Thanks!

Dave

[Jon Paul]

Hi Dave,

Lots of questions here, I'll try to answer them all.

Any performance comparisons between 2000s and 1500s?
Short Answer: no

1500s vs 2000s performance equal?
Short Answer: No
There are a myriad of reasons that the 1500s is slower than the 2000s, including controller, controller cache, drive architecture, and drive spindle speeds. The 1500s is targeted at a small workgroup the 2000s is targeted to a mid-sized workgroup.

Does adding RAM help?
Short Answer: It depends
I'm not prepared to open that can of worms. When trying to determine your best bang for your buck in improving performance, use the Performance Monitor in the OS. Also there are a number of benchmark programs available to test from network thru to backend storage. (IOZone, IOMeter, NetBench)

[Steven Clementi]

You might want to keep in mind that SATA drives are alot more likely to fail in a 24/7 environment.

MTBF(MTBU) for scsi is measured at 80% utilization(duty cycle)

For SATA, it is measured at 20%.


As for performance, it's hard to say. Depends on Controller... controller cache, system memory, etc.. lots of factors. Since the SATA drives are 7200RPM and the SCSI are probably 10KRPM... that would make a difference too.


Steven