WD2010 Benchmarks - Inconclusive Step Rate Results

Peter Fales psfales at cbnewsc.ATT.COM
Tue Aug 29 13:52:03 AEST 1989


Well, I tried out Lenny's handy disk exerciser.  (It really gave that
poor disk a work out - the table was actually rocking back and forth).
My results are still puzzling - I can't get the disk to speed up, but
I can't get it to slow down either.

The procedure I used was:

1) Change the step rate in the VHB using iv -u
2) Reboot
3) Verify the change with iv -t
4) Run the "testit" script which does a series of long seeks, then
	a series of random seeks.
5) Grep for "Average" in the results file which prints the average
	for the long seeks, and the average for the random seeks.
	
My disk is a Miniscribe 6085:  67MB, 1024 sectors, 8 heads, 28 msec.

Step rate 0:

long_rand0:Average seek time: 9
long_rand0:Average seek time: 5

Step rate 14:  (Should be faster?)

long_rand14:Average seek time: 9
long_rand14:Average seek time: 5

Step rate 13:  (Should be MUCH slower)

long_rand13:Average seek time: 9
long_rand13:Average seek time: 4  (!!)

The last results  have me still thinking that something is wrong 
with the procedure - the software is somehow not using the step rate
parameter from the VHB, but I don't know why not.  

-- 
Peter Fales			AT&T, Room 5B-420
				2000 N. Naperville Rd.
UUCP:	...att!peter.fales	Naperville, IL 60566
Domain: peter.fales at att.com	work:	(312) 979-8031



More information about the Unix-pc.general mailing list