#Hard Drive Advice
22 messages in this thread
Well, I finally got around to the hard drive issue again. I installed the
termination resistors on the Quantum and then made sure it was installed on the
last connector on the cable. Then I fired up the system and, of course, I had
to warm boot because I got the plain cli with no command recognition
again…started up ExpertPrep and on page 2 set the Quantum to mount instead of
boot, then on page 1 used the option to write information to hard drive (or
some similar wording). I then turned off the power and allowed all rotating
machinery to spin down. Turned it back on and voila!…complete,
uninterrupted, error-free coldstart using the correct hard drive (Maxtor), and
it seems that both drives are operating properly at this point so I see no need
to reprep or reformat the Quantum. Perhaps at some time in the future I will
rearrange things so that the Quantum is the boot drive since SysInfo 3.23 tells
me that the Quantum's data transfer rate is over 2x that of the Maxtor
(2,080,507 bytes/sec vs. 987,979 bytes/sec).
Thank you for all the helpful information. The messages I received from you
and Trent contained so much useful information that I have kept them for future
reference.
//daniel on autopilot in paradise
\X/ powered by intel and motorola
daniel,
It sounds to me that maybe I should reprep these drives here to the mfgr's.
specs instead of the ExpertPrep setup. I'm not getting any where near 2,080,507
bytes/sec on my Quantum, more like 1,575,000. I think that extra 1/2 MB per
second is significant.
Glad your now up and running OK.
Christopher
I didn't remember the speed difference to be that much. I guess I'll just
leave it alone with the mfr's specs.
Is your machine accelerated? That may (probably) affects data transfer rate.
//daniel on
\X/ AutoPilot 1.75
daniel,
Yes, my A2000 has an 33Mhz '030 and 4MB 32 bit ram and 4MB of 16bit ram besides
the 2MB of chip and the ECS chipset. Hmmm…maybe having the mix of 32 & 16 bit
memory causes things to slow down on SCSI reads/writes? I don't know, it does
seem slower than what you report with SysInfo. I've the same version you
reported using to check your disk speed.
Christopher
Chris,
Do you remember whether it is possible to use both the 2090A and the 2091
controller in the same computer? I seem to remember that, if slots were
available, you could use both at the same time. Am I right?
—Betty
Betty,
Sorry, I never have owned or used either of the C= controllers.
I'm sure someone else knows and will provide the answer.
🙂
Chris
Thanks anyway.
—Betty
Yours is definitely accelerated well beyond mine so I'm sure that the data
transfer rate is more a function of the parameters (cyls, hds, spt, etc.) used
to set up the drive, and I don't necessarily think that faster is better
because there are other factors to consider such as storage capacity,
read/write errors (data integrity).
I checked this one several times with SysInfo and got consistent results, so I
don't know what else could be going on.
//daniel on
\X/ AutoPilot 1.75
daniel,
I've checked the drives a couple times with that SYSINFO and I got consistent
readings of 1,539,005 bytes transfer rate. I haven't any other way to confirm
those results. The Maxtor 345MB drive is getting a bit less, aroun 1,518,208. I
think it is this Series II controller. It seemed that I had faster results
using a CLtd KRONOS controller. It always seemed faster. Need to hook up one of
these new drives to it and test. Only thing is the KRONOS doesn't support the
rigid disk block, but that is alright.
Christopher
Could the difference in transfer rates be a result of the difference in the
amount of data stored on the drives? (yours is a LOT fuller than mine right
now)
//daniel on
\X/ AutoPilot 1.75
Chris,
Is your controller a GVP Series II? I ask as I run the same so it may not be
that. Then again it may just be. SCSI is black magic as far as I am concerned
and am happy if the stuff works without a problem <g>.
Trent
'Dulce et decorum est pro patria mori'
Trent,
Yup, it is indeed a GVP Series II controller. I have a KRONOS in the box too,
though just the scanner and a printer connected to it.
Christopher
Chris,
Drive speed has little to do with the chipsets or memory. It's all in the
drive. Usually the higher grade Quantums (ie LPS series etc) are twice as fast
as the lower grade (the ELS series). I have two Quantums on my system, one of
each grade. I get the following numbers
DH0: Quantum ELS85S 990,468 bytes/sec DH1: Quantum LPS240S
GM240S01X 1,913,459 bytes/sec
The 240MB HD has a faster spin rate but it's almost silent as compared to the
85MB HD (the noise from the little beast almost drives me to distraction some
nights <g>). Costs a little more, but worth it in my estimation.
Trent
'Dulce et decorum est pro patria mori'
Trent,
Drive Model bytes/sec
===================================
Quantum Lightning LT540s 1,539,005
Maxtor 7345SR 1,518,208
No where near the 2MB daniel reported getting.
Christopher
Chris,
Just for giggles, I reconfigured the Quantum last night:
LowCyl: 1 BootPri: -10 DMA Mask: 0xfffffffe
HiCyl: 5782 Partitions: 1 MaxTransfer: 0x7fffffff
Heads: 3 FileSystem: FFS Boot/Mount: Mount
Sectors: 62 BufferType: Either BufferMemory: 32
MB Size: 525
Several SysInfo speed checks gave the following results:
Model bytes/sec ================================= QULT540S 2,075,018
to 2,091,574 7120SCS 984,270 to 985,503 DF0: 22,206
The Maxtor is 89% full whereas the Quantum is again empty except for trashcan,
disk.info and mountlist. I think transfer rate is affected by volume of data
stored on the drive, but I don't have time right now to copy several megabytes
of files to the Quantum to fill it and then test it… maybe another time.
//daniel on
\X/ AutoPilot 1.75
Chris,
Yeah. My best (the 240LPS) was approx 1.9MB. Hmmm.. Originally the LPS was
to go into a Mac (the setup software that came with the drive was for a Mac).
Wonder if that could be a reason for the differences is speed? Or would it
really make a difference?
Trent
'Dulce et decorum est pro patria mori'
I phoned Quantum Tech Support today, er, uh, yesterday. The QULT540S has 1049
cyls, 16 hds and 63 spt but since it's SCSI, the firmware response when queried
such as by ExpertPrep is perfectly acceptable. In fact, the engineer indicated
that as long as the combination of parameters used to set up the drive does not
exceed the capacity of the drive it should function properly, and without
significant variance in the data transfer rate.
Just thought you'd wanna know. It's late so I'm off to experiment with the
Quantum before hittin' the hay.
//daniel on
\X/ AutoPilot 1.75
daniel,
I kinda thought that Fastprep was reading the firmware correctly. I remember
hearing a long time ago something similar to what that engineer told you.
I'll try and experiment with my second 540S and see how much different it acts
after being reprepped with the new info.
Thanks,
Christopher
My pleasure Daniel. Quantum is the way to go (of course _all_ of my drives are
Quantum <g>
Trent
'Dulce et decorum est pro patria mori'
I agree with you for SCSI… when I ordered my GVP Series II with 120M HD, I
expected to get it with a Quantum. You can imagine my disappointment when I
opened the package and saw a Maxtor drive on the card. I have no complaints,
however, since the Maxtor has given me absolutely no problems whatsoever. I
do, however, wish that it had a bit faster data transfer rate but I suppose the
Quantum gets it by the faster spin.
//daniel on
\X/ AutoPilot 1.75
I called Quantum yesterday: the QULT540S has 1049 cyls, 16 hds, 63 spt but the
engineer said that since it's a SCSI drive and the exact parameters used to set
up the drive could vary as long as the capacity of the drive is not exceeded.
He stated that the default that ExpertPrep reports would be fine since it's a
result of the firmware query.
So, off to experiment….
//daniel on
\X/ AutoPilot 1.75
=>that ExpertPrep reports would be fine since it's a result of the firmware
=>query.
That's my understanding as well. Expertprep gets the figures it does from
querying the drive itself. The numbers it then displays are the factory
settings. I would think that they would be the way to go (why put wonked
numbers in the firmware otherwise). Anyways, have fun.. <g>
Trent
'Dulce et decorum est pro patria mori'