Hi friends:

I'm so fed up I could just scream. My main workstation drive-a Samsung SATA SP1213C,
just died a couple of weeks ago. I had only been in use for about nine months. After that fiasco,
I decided not to trust the replacement refurbished drive, and go out and buy a new one.
I thought that would be safer. What a mistake that was.

I did lots of research, and decided to go back to the old reliable SCSI standard I used to use.
So I paid a lot of money for a brand new "enterprise-grade" Seagate ST373207LW 10K.7 U320 73GB drive.
It's sitting on an Adaptec 2940UW controller. I also have another SCSI drive on that chain-
a Sun (Fujitsu) MAG3091LC 9.1GB 10K RPM drive.

Now what happens? I get disk problems on the phracking new drive. It's bizarre-each time
I go to check the System Event Log, it gets corrupted after a minute or two of browsing through entries.
Then, when I do a chkdsk, it either reports having to repair index entires, or that the volume bitmap
is corrupted and needs to be repaired.

When this first started, I originally had an error entry in the System event log for the new drive.
I ran chkdsk, it reported some minor error, and then it supposedly fixed the NTFS filesystem.
I thought the problem was solved. Again, I was so wrong.

Now the errors are back. They seem to show up when I view the event logs. I don't know if
viewing the System Event Log causes the corruption (some kind of software bug) or that the
System Event Log keeps getting corrupted because it's on an area of the disk which might be going bad.

But for Dang sake, I just go this drive not even a week ago, and it was supposed to be
a premium product. The catch-22 is that each time the corruption takes place, I can't read the
System event log, so I have no error message to go on.

I'm pretty sure my SCSI cabling and termination are correct. I've run memtest for 12 hours
on the RAM and it passed numerous times.

I've checked the voltages on my power supply. They're not perfect, but to me, they're well within
reasonable range-->as follows:
12V-->12.22V
5V-->5.10V
3.3V-->3.34V

IS there something else that could be going on with the power supply?
It's an Enermax Liberty ELT-400 watt model.

According to the SMART data, as reported by SpeedFan

ECC ON-the-fly hardware corrections counter: 847175
Total Posted Recoverable Read Errors Counter: 847175
Verify errors without substantial delay counter: 1337160

Update: IN order to preserve the System event log file, I copied
it to my external USB
hard drive. When i went to open it with the Event Log Viewer, the Viewer reported that it could
not open the file as it was corrupted. WTF? I'm totally confused at this point.

I'm so fed up with disk drives at this point, I just want to throw the damn thing out the window.

Could the problem still be power-related?
Could it be a termination problem?
Wrong cluster size specified?

Again, I'll recheck the cabling/termination....but I'm pretty sure they're
correct.

Any help at this point would be appreciated, as I'm losing my mind at this point-the symptoms are all so contradictory.

Component list:

Abit KN8-SLI (Nforce4)
Adaptec 2940UW SCSI adapter
Seagate 373207LW 10K.7 U320 73GB drive
Sun (Fujitsu) MAG3091LC 10KRPM U160? 9.1GB drive
NEC ND-3550a DVDRW (IDE)
2x512MB Geil RAM
Enermax Liberty ELT-400AWT power supply


Thanks for any suggestions.


Shplad

Hey men,
I do not know what do you want, data on the dead drive or what?

hddlab:

Thanks for your reply. I would've thought statements like these made it
pretty obvious:

I can't read the System event log, so I have no error
message to go on.
.
.
Is there something else that could be going on with the power supply?
.
.
When i went to open it with the Event Log Viewer, the Viewer reported that
it could not open the file as it was corrupted. WTF? I'm totally confused at
this point.
.
.
Could the problem still be power-related?
Could it be a termination problem?
[Could it be] Wrong cluster size specified?
.
.
Again, I'll recheck the cabling/termination....but I'm pretty sure they're
correct.
.
.
Any help at this point would be appreciated, as I'm losing my mind at this
point-the symptoms are all so contradictory.

I'm looking for diagnostic help, thanks. The data is already rescued. I'm
not even sure it's in serious danger. Yet.


Thanks.


Shplad

Well, I don't for sure which fixed it, but a friend of mine remembered
that the Adaptec SCSI card on which I had the drive running had some
firmware updates released way back. One of the issues listed in the
changelog was "read/write errors".

I flashed the Adaptec card to the new firmware, did a low-level format of
the drive, and reinstalled Windows.

Just in case, I didn't install some of the other software that might have
kernel mode access which might cause disk corruption, such as the newer
version of the UPS software I recently switched to.

So far, so good. We'll see what happens.


Shplad

Okay, here's an update, in case anyone uses SCSI and/or
these drives:

It's starting to look like the problem was not the new Seagate
drive itself, but having the Seagate drive, and my old drive (Fujitsu
MAG3091LC U2W) on the SCSI drive chain at the same time.
As soon as I added the Fujitsu drive back into the chain, the errors
started occurring again. As soon as I removed the Fujitsu drive,
the errors began to go away again.

Now, let me make this clear: The Fujistu drive is in good health.
Interestingly, no data corruption was found at any time on it. It appears
possible that having the two drives, of different speeds, may confuse
communication on the SCSI chain. I'm not really sure yet.

Can anyone point me to any resources or other information on the
subject of data corruption when mixing drives of diff. speeds/types
on a SCSI chain?

For three days now, as long as the Seagate ST373207LW is the
only drive on the SCSI chain, there has been no data corruption,
and chkdsk reports good health. For that reason, I will only
be using the Fujitsu as a spare in emergencies.

So much for the compatility the manufacturers claimed between
drive standards. At least for now, it looks combining these two drives
on an Adaptec 2940UW, at least on my machine may cause data corruption.


I hope this helps someone.


Shplad

Forget everything I said-the drive corruption is back. The event log is
no longer corrupted. That turned out to be a separate, different Windows
bug for which Microsoft has a workaround.

BUT... the main symptom with chkdsk reporting either volume bitmap errors,
or indexes needing repair, is still there.

I called Seagate. Their 1st level support guy seemed helpful. We triple-checked
cabling, termination etc. He advised me to ensure I was using a U320 compliant cable
and active or multimode terminator. I even tried another SCSI cable. He repeatedly
assured me that this drive was definitely compatible with the 2940UW
card. Eventually, the same problems came up again.

Then I spoke with a 2nd level support team leader at Seagate, who relayed my message
onto a 3rd level support rep. Here was that 3rd level rep's brilliant advice:



So basically, someone who is supposed to be a 3rd level drive specialist is recommending
I leave a scsi chain unterminated. Whah? Thanks Seagate for your total incompetence.
I'm pretty sure he didn't even bother to read my case history.

Oh, and the story gets better-I ran the Linux-based command line "ST" utility from a Linux DVD
from which I booted. It hosed something on the drive so that the machine wouldn't even boot
at all. I tried repairing the MBR, the Partition boot record-nothing. So I had to reinstall Windows
again from scratch.

I put in a phone call to the 2nd level team leader, and he hasn't even bothered to return my call.
The is the level of support I get for buying a premium-priced
enterprise-grade drive.

My last hope is that I'm going to try using a U160 or U320 controller.

But I must say, I'm appalled at the disgusting way Seagate has handled,
or should I say "not handled" my case.


SHAME ON YOU SEAGATE. My drive problems may or may not be caused by a U320 drive
being connected to a 2940UW. But if they were, it's not my fault. I even phoned Seagate tech.
support before I bought the drive to verify it would work on my card. They assured me repeatedly
that it would work just fine.

I think they misled/lied to me. Big time.

I'm still pondering whether I should get a used Adaptec U160 LVD controller, or just return the drive and go back to IDE.


Shplad