#Hard drive baffler
22 messages in this thread
I just bought a 1.06 gig Seagate Fast SCSI2 drive, but I can't seem to format
or partition it. It's model ST31230N, if that means anything to you.
HDToolBox fails to correctly read the number of cylinders, heads, blocks per
track, or blocks per cylinder, while the version of HDToolBox that came with my
WarpEngine gets the heads and cylinders right. Both versions correctly read my
Maxtor SCS1 drive.
HDToolBox told me that the number of heads times the number of blocks per track
should equal the number of blocks per cylinder, which makes sense I guess, so I
played with numbers and arrived at
3992 cylinders (according to the drive manual and the HDToolBox-WarpDrive)
5 heads (according to the drive manual and HDToolBox-WarpDrive)
109 blocks per track (which gave me 1062MB, which I assume is correct)
545 blocks per cylinder (following the formula)
When I try to low-level format the drive, which is what my Amiga Hard Drive
manual and my WarpEngine manual tell me to do, the drive light goes on, but
there's no activity – I can't hear the heads moving, and nothing happens.
When I try to Save Changes To Drive, I get Error 20, can't write to drive,
whatever that means.
The drive has jumpers for termination and parity, among others. I left the
drive at ID 0, which isn't in conflict with anything else, and tried with and
without parity checking, which made no difference.
Does anyone have any idea what's going on? I'm baffled. I'd GREATLY
appreciate any help. TIA
Mr. Zombie
Paul,
as far as I know you should never try to low-level format a SCSI hard disk.
Dunno why, really, it just says so everywhere I look. SCSI hard disks are
always low-level formatted straight outta the box. Apparently most SCSI hard
disks even ignore the low-level format command and just return ok or so. You
just run the HDToolbox equivalent which came with your SCSI host adapter,
usually just accept the default values it shows you and exit with saving. This
will write the settings into the disk's RDB (rigid disk block) after which you
only have to DOS format them either from the Workbench's menu: 'Icons/Format
disk', or from the shell: 'format device|drive <device name> (for example
'DH0') name <name> [ofs|ffs] [international|nointernational]
[dircache|nodircache] [noicons] [quick]'. '|' delimits choices, […] surrounds
optional parameters, <…> surrounds user provided data. After that you should
be able to use it right away (might be after a reboot only).
// Christian "ELP" Busse
\X/ near Frankfurt, Germany
The thing is, though, that whenever I tried to save information to the drive,
both versions of HDToolBox told me Error 20 – Can't Write to Drive.
I even completely broke down my machine, took out the WarpEngine and every
other card, and tried to do it with the motherboard controller, and got the
same results.
Mr. Zombie
Paul,
My experience is that HDTools is the LAST resort when it comes to SCSI
stuff. Your host adapter (GVP?) should have come with something much
better than HDTools.
Is the Maxtor still in the chain?
Is the Seagate new or used?
Where's the termination?
What host adapter are you using?
Gary
on AutoPilot from Charlotte, NC
I'm using my WarpEngine's FAST SCSI-2 controller. It comes with a version of
HDToolBox for the WarpEngine, which I tried, and a program called ScsiTools,
which I tried. I also tried C='s HDToolBox. All three gave me the same
results – they failed to write to the drive, and if I tried to Initialize the
drive – ScsiTool's lingo – or Low Level Format it – HDToolBox's – the drive
light would go on, but there would be no drive activity, and that would be the
end of that.
I also tried the drive on my 3000's motherboard SCSI controller, with the
WarpEngine removed from the machine. Same results.
I tried the drive on the middle connector on the internal ribbon cable with the
Maxtor still up and running, and on the end of the ribbon cable with the Maxtor
removed. When the Maxtor was in, I disabled termination on the Seagate; when
it was out, I left the Seagate in its factory default termination settings –
termination on, termination power supplied by the drive. I also tried
termination power provided by the SCSI bus. All gave me the same results.
The Seagate is brand spanking new.
I ran a pd utility called ProbeSCSI which I got here in the libs. It reported
that the drive actually has 104 blocks per track, not the 109 I had guessed. I
changed the Drive Type values to reflect this, but this did not solve the
problem. ProbeSCSI also told me that there was a read error off the drive.
Most of its output, though, was Greek to me.
Does this give you any idea what's going on? I appreciate your time and
efforts. Thank you.
Mr. Zombie
Hi Paul,
I once had similar problems with a Seagate 2G Barracuda drive (forgot the model
number).
I couldn't write to the drive like you.
Solution was to low-level format the drive (it took very long).
Note: Not all programs really do a SCSI-low level format!
E.g. the HDToolBox does NOT SCSI format the drive if you select LowLevelFormat!
I used my own EBprep for this job, but the HDInstTools from the Oktagon
controller will do so, too.
As I have now released a new Oktagon HDInstTools disk version (3.4) I will put
it here in the library.
It is not only of interest for Oktagon users as it should work with other SCSI
host adapters, too.
Older versions refused to work with non-AlfaData/bsc controllers.
Best regards,
Oliver
Thanks Oliver, I appreciate it. Btw, a friend of mine is trying to find an
Oktagon for his 2000 but has had absolutely no luck whatsoever. Are they still
being made?
Mr. Zombie
Hi Paul,
>> Thanks Oliver, I appreciate it. Btw, a friend of mine is trying to find an
Oktagon for his 2000 but has had absolutely no luck whatsoever. Are they still
being made?
Yes!
Oliver
Oliver-
Could you recommend a place to find one, then?
Btw, I d/led your SCSI utils, but was unable to find a way to make it look at
anything other than scsi.device. Is there a tooltype or something for that?
Mr. Zombie
Hi Paul,
>> Could you recommend a place to find one, then?
You are in the USA, I suppose? Well. I'm not! <g>
Sorry, no idea. Only thing I know is that they are produced and they are
available.
Ask your dealer for it.
>> Btw, I d/led your SCSI utils, but was unable to find a way to make it look
at anything other than scsi.device. Is there a tooltype or something for that?
Oops, sorry! This is undocumented, of course… <g>
The Tooltype is simply DEVICE=scsi.device
From the CLI type HDInstTools scsi without ".device"
(I've done this, because I am a lazy typer)
Oliver
If you have not tried already, set reselection to OFF using BattMem. The
Maxtor will definately respond honestly, truthfully and correctly to a reselect
request — a GREAT many other drives do not do so.
Both of the symptoms you described look like reselection based errors.
-bill
Wow…you really have tried everything!!
I've lost your previous letter. What was the Seagate model number? And do I
remember correctly that you did try DOS formatting? (the other writer was
right, SCSI drives are low level formatted at factory…and sometimes you can
screw them up if you try to re-do it.) You should have been able to go directly
to ADos formatting/partitioning. Send me the model # again, so I can check some
data I have here. But if you want a gut reaction, given everything you've
tried, you just might have a "bum out of the box" new drive. It happens.
Gary
on AutoPilot from Charlotte, NC
It's a Seagate ST31230N. I tried the button called Low Level Format in
HDToolBox (both versions) and Initialize in ScsiTools, and neither worked. The
drive didn't appear on my Workbench, so I couldn't try the format command. I
was able to define partitions based on the blocks per track and so on info I
entered in Change Drive Type, but I wasn't able to save ANYTHING to the drive.
I probably do have a bum drive, because I tried it on 2 different controllers
with 3 different programs as well as ProbeSCSI, which reported read errors, but
if you have any more info, HIT ME! <g>
Mr. Zombie
Paul,
According to the docs I have downloaded from Seagate, there is a write protect
jumper on this drive!! It's hard for me to decipher exactly, since this is
ascii text, and diagrams don't do real well, but it appears to be on the side
of the drive, right where the termination power selecters/jumpers are and
appears to be jumper #8 (the last one, closest to the front of the drive) in
that group. That could certainly account for your problem! Try changing that
#8 jumper around to the opposite of what it is now and see what happens.
Gary
on AutoPilot from Charlotte, NC
Gary-
I think you're diagram is for a different drive. There is in fact a Write
Protect jumper, and it is in a series of 8 on left bottom of the drive, the
power connector side, not the ribbon connector side, but the one nearest to the
front of the drive is labeled TE, for Termination Enabled, while WP, Write
Protect, is the 5th from the front, the 4th from the power connector. The
manual indicates that the default state, unjumped, indicates NO write
protection.
I admit I didn't try it both ways, and I should have, but this wouldn't explain
the difficulty reading the drive's info. Still, I'll try it.
And thank you for the suggestions. In a sea of emptiness…
Mr. Zombie
HDToolBox in version 3.x is a very good program.
-bill
Paul,
Check your jumpers. Some drives come with write-portect jumpers for use in
like BBS libs or controlled databases. Maybe you got a jumper where it
shouldn't be. Or worse, your cable is upside down:^>
wmc – via Autopilot!
Wayne-
I did double check that. Unless the manual is wrong, it's only write protected
when there's a jumper on the WP location, and there isn't. But what ProbeSCSI
reported is a read error. I'm wondering if the drive isn't defective.
Mr. Zombie
Sounds like it might be time to test the warranty policies of Seagate. I know
it is near impossible to talk to a live human at tech support, but they do have
a forum hear (GO SEAGATE) where you can get quick responses and perhaps even an
RMA. Personally, I like Micropolis and Quantum better. Call the toll free
number, tell the human your problem, get a cross ship RMA/parts order have the
replacement/parts at your door next day.
wmc – via Autopilot!
Paul:
Did you check the 'WP' jumper on the drive? Could the drive be Write
Protected? (A very slim chance, but who knows?!?)
(Actually, in reading your later posts, I see you've already checked.
It sounds more & more like the drive might be toast? It _can_ happen….)
Erik Flom