Hardware problem? Software Problem? HELP!

I have a Dell Inspiron 5150 laptop (3.06 GHz / 2 GB (maxed out) / 250 GB / 1600x1200 15" screen). This model was widely known to overheat, and as a result of a class-action suit, Dell replaced the mobo about 3 years ago. It has run flawlessly the six years I've had it.

I recently upgraded the 60 GB hitachi drive that came with it to a 250 GB WD drive. After a while, it said it wouldn't boot. I booted to Knoppix and everything was still there. I copied what I needed off to a USB drive, wiped the drive clean, and started over. Everything was great for a week or so, and then it did it again. I put the old (HitachI) drive back in, and it booted a time or two, and then it apparently died of old age. it HAD been just fine, but now it only clicks, and the Hitachi diagnostics say it's hosed. I assume this was a coincidence, but maybe not.

So, back to the (new) WD drive. I ran WD's diagnostics on the drive, and Dell's on the laptop, and everything passes its diagnostics. I zeroed the drive, re-installed XP Pro, loaded SP3 and all of the associated updates, re-installed my apps, and everything was fine. For about a week.

A few days ago, it again popped up "A disk read error has occurred. Press Ctrl+Alt+Del to restart," which just takes me back to the BIOS. I see the message of this text in the boot sector of the first partition, so I know it's coming from there.

Knoppix, booted from a CD, can mount the drive just fine, so I know it's not the partition table, and the (NTFS) file system is obviously intact. I had anti-virus software on it (don't recall which), but it never complained about a boot-sector or MBR virus.

Given that it's six years old, I suppose it COULD just be dying of old age, but I'm reluctant to think so, and even more reluctant to part with lots of cash to get a new one, if I can solve the problem with the old one. If it's a hardware problem, how come it keeps showing up the same way -- by leaving the hard drive unbootable?

I tried the Windows XP Boot floppies (took about 15 minutes to get through all six), hoping I could get to the Windows Recovery Console. On the sixth floppy, it blue-screened with something to the effect of "BOOT_VOLUME_UNMOUNTABLE". So, no recovery console for me.

Any suggestions?

 

Have you tried Western Digital Drive tools?

http://support.wdc.com/

 

Pardon my ignorance -- when you say their "drive tools," the only thing see is the "Data Lifeguard Diagnostics". Is that what you mean? I downloaded the floppy version. THAT's what I was referring to when I said I had run the WD diagnostics. Is there something else I should look for on the WD site?

The drive is a 250 GB EIDE (PATA) drive, so I went to the link you posted, clicked on "Downloads," "WD Scorpio", "Scorpio Blue (EIDE)" and my model is "WD2500BEVE", so I assume that the only tools I can download are TrueImage (which I believe is a Ghost-ish sort of thing) or the diagnostics (which I *did* use).

Did I miss something?

 

Western Digital produce a number of versions of their Data Lifeguard for different situations - SCSI - SATA -IDE -External -Internal - DOS from bootable floppy.

Their site has lots of information and FAQs about their drives.

Otherwise a good compilation is at

http://www.asci-red.net/tools.htm#hmt

 

Could it be a delayed-write problem? The new drive DOES have a significantly larger buffer (8 MB) than the old one. Isn't one of the last things Windows does at shutdown is to set a "dirty" flag? I wonder where that is stored. Just a thought. I'm stumped.

BTW, MHDD says the drive is fine. It's looking more and more like it's not a hardware problem, but I can'r rule anything out just yet.

 

http://support.microsoft.com/kb/555302

 

Thanks, studiot, but I don't think that's it, either.

That page suggests that the cause can be:

1. The file system is damaged and cannot be mounted.
-. This isn't the case, because Knoppix CAN mount and read it.

2. It has a 40-vs-80 wire cable.
-. It has the same cable it's always had. The Hitcachi drive supports ATA-6 (UDMA): http://www.hitachigst.com/tech/techlib.nsf/techdocs/85CC1FF9F3F11FE187256C4F0052E6B6/$file/80GNSpec2.0.pdf (Page 9), so I assume Dell surely used an appropriate cable to support the spec. I suppose it's possible that the connector has suddenly gone bad, but this seems a bit unlikely -- the WD diagnostics have read every sector on the drive a few times; you would think that an intermittent connection would have resulted in at least one read failure.

3. BIOS is configured to force a faster UDMA mode.
-. There's no setting in the BIOS to configure the interface to the drive. The BIOS *DOES* show the drive as being 137 GB, but this limitation did not seem to bother XP. When I installed it, XP saw the full 250 GB, formatted it just fine, and wrote data to MORE than 137 GB of it before it checked out.

I'm leaning more and more towards a software problem.

thanks for keeping those ideas coming, however!

 

Which service pack was the installation CD? If it was an original, no SP, it may be at the root of your problem as it does not play well with large disks.

This would indicate that the BIOS is not currently supporting 48-bit LBA, you need to check for a later BIOS version.

 

I was installing from XP Pro SP2. It DID install, and there's not a newer BIOS for this laptop.

I've been doing some research, and I *have* found lots of folks that say that you can't go over 137 GB in this machine (I understand the whole LBA issue). As I understand it, though, if the O/S supports it, then what the BIOS says doesn't matter. Is that the case?

Knoppix seems to have no problem with this drive, so I assume it's not an inherent hardware limitation. If it's a software limitation, then it seems I should be able to get Win XP SP2 to go back on it. I *had* Win XP Pro SP2 on it, and it worked fine for a week (or so) -- twice. If there's some inherent limitation, then it seems it shouldn't have worked the first time.

I do happen to have an old 40 GB spare drive sitting around. I suppose I can put THAT in, and see if I can load XP on it. If so, then I'll probably try to just buy a 120 GB drive, and leave the 250 as a spare for some OTHER project, or sell it.

Has anyone had any experinece with getting around a BIOS-imposed limit of 137 GB, or am I completely off on a tangent?

I really MUST have a reliable, working laptop in the next three weeks, and I really HATE the idea of buying a new one, if this one can be coaxed into working (properly) again!

 

There's a poster on the Dell community forums who says the 5150 BIOS doesn't support 160Gb drives and suggests using a 120Gb.

 

Post a copy of boot.ini (rename it boot.txt)

 

Satrow,

I saw that post. I've been scouring the net to see if anyone has comments on this. Dell's "spare parts and upgrages" for this model doesn't offer anything bigger than 80 GB, hinting at compatibility issues. I've also seen where someone said they were using a 160, but had partitioned it for 120.

In my experience, "doesn't support" can mean three things:

1. It simply doesn't work, so don't bother ("round holes don't support square pegs")
2. Sometimes it works; sometimes it doesn't, and if it happens to work for you, that's great; if it happens to not work for you, we won't help you MAKE it work.
3. It seems to always work, even though we didn't design it that way. But BECAUSE we didn't design it that way, we're not guaranteeing that it will work.

Since (1) Knoppix doesn't have any problem with the 250 GB drive, and (2) Windows DID install and work fine for a week or so, I don't think it's #1. I have to remind myself, however, that Knopix and XP are two completely different beasts. The fact that Knoppix works with the drive tells me that it's not some fundamental hardware incompatibility.

Since I'm having the trouble I'm having, it would seem to be #2 or #3.

Looks more and more like #2. Today, I'm going to take the 250 GB drive, shrink the existing 250 partition (mostly empty) to something less than 120, and partition it into 120 and 120, and then see if it boots. SURELY once XP is up and running, it will be able to handle the second partition for data, even if it can't handle BOOTING from a partition > 128 GB. If it doesn't, I can always live in the first 120 GB, or I can try installing it on a spare 40 GB drive and see how that goes. I can't live with just 40 GB ongoing, but if itworks, then I can buy an 80 or 120.

As I recall (a lot has happened since then), I originally took the 60 GB drive (which was working fine at the time), and copied the boot partition as an image to an external drive, swapped in the 250 GB drive, formatted it as 250, copied BACK the partition, and then used ntfsresize under Knoppix to make it fill the 250 partition. It booted on the first shot. I keep getting hung up there -- it DID work for a while. I probably should have backed up the MBR and the first few sectors of the boot partition, so I could compare before/after, but c'est la vie.

Studiot, Here's boot.ini:

[boot loader]
timeout=30
default=multi(0)disk(0)rdisk(0)partition(1)\WINDOWS
[operating systems]
multi(0)disk(0)rdisk(0)partition(1)\WINDOWS="Microsoft Windows XP Professional" /noexecute=optin /fastdetect

 

Are you sure that XP is on partition 1 ?

What with all the fiddling about with Linux and such.

Depending how you cloned the 60G drive. Dell starts the disk with a small partition.

What is the reported size of the partition on the 250.

I have just fully formatted a WD 250G drive to yield a single partition of 232G, covering the entire disk.

Also redetect the drive in Bios.

 

Basically with a 48-bit LBA compatibility problem, XP usually works fine until it hits an error with the file system/partition then it's U/S until you zero the disk and start over. Partitioning down to 2x120Gb or thereabouts - first partition should work 100%, I'm not sure about whether I'd trust partition 2 with any valuable info. You could try to use drive overlay software (from your HDD maker?), that should solve the problem but it does add another layer of complexity should something go wrong.

If you need simplicity and stability, go for a 120Gb drive.

 

Satrow: What's "U/S" ?

Studiot: I'm sure that's the first partition -- I didn't clone the whole drive; just my main partition. Dell had a 32 MB partition for diagnostics (type 0xDE). When I zeroed the drive, I blew it away. Now there's a single primary parttion (boot flag set, type 07), starting in sector 63, running through sector 488,375,999. At 512 bytes/sector, that's 250,048,511,488 bytes. If we take a gigabyte to be 2^30, rather than 10^9, that's 232.875 GB

 

Update: It sure looks promising! Here are the details, in hopes it will help someone else...

I couldn't get Knoppix's ntfsresize program to resize my 250 GB partition downto 120 GM, even though it was only using about 25 GB. Something about "(1032>1024)". I was using the Knoppix 5.1.1 CD, which has an older version of the ntfsprogs, so I got the 6.0 CD (which uses the new 2.0.0 version of ntfsprogs). THAT wouldn't resize the partition, either - same message. I looked up the particular message, and others had suggested not trying to shrink the partition so much, but I was going from 250 at 10% full down to 120 at about 25% full -- that should have left PLENTY of working room. There's apparently something else keeping it from being able to resize the partition. Maybe it's the 2 GB swap file in the root. Who knows. Whatever.

Some digging around on the web indicated that Mandriva's resizer might work. I downloaded Mandriva One 2009 (Spring), burned the CD, and couldn't get it to boot. Eventually, I figured out that I had to disable ACPI. Once it booted, I opened a command window, and ran "diskdrake" (Mandriva's disk manager / graphical partition manager / resizer). I used it to resize the 250 GB partition down to 120, and it did so without a single complaint. I then created ANOTHER partition (NTFS-3G) in the remaining space. I told Mandriva to restart.

When it restarted, Win XP Pro came up, absolutely intact, on the first shot! It wanted to chkdsk the drive (no surprise; even ntfsresize flags the drive for a chkdsk on next boot). Chkdsk made two corrections to the index on the file system, and then it rebooted again, and came right into Windows.

This time, it said it found new hardware and installed the drivers. I'm not sure what it thought the new hardware was, but it was probably the new partition, so it loaded disk drivers. At any rate, I let it reboot, and all was still fine with the restart.

My apps SEEM to be all there (I started a couple without incident), but I haven't done any extensive testing (yet). I updated on my antivirus definitions, and downloaded and installed a couple of app updates.

Next up -- I'll try putting some non-critical data on the data partition, and use the system for a few days, manipulate that data, bring it up and shut it down, and see what I can determine from there.

Satrow and Studiot, thanks for your help and suggestions. I'll post an update in a week or so, and document how things have gone, with respect to both the boot partition AND the data partition. If it gives itself another lobotomy, I'll probably be back sooner!

Thanks again,

Larry Thomas aka 'ThomasLG'

 

Keep at it Larry.

I have never seen any problem with larger disks, partitioned and formatted by another system.

I expect you will be OK now.

How about posting the total size ? I take it there was no small Dell partition.

 

Thanks, Studiot.

C: is 120.00 GB
D: is 112.88 GB

So, it looks like I got my 232.875 GB, for 250 billion bytes. I wish the drive manufacturers would use 2^30, rather than 10^9 for "Giga"!

I deleted the Dell partition. All that was in it was diagnostics (it was only 32 MB), and I can load them from floppy if I need to.

 

U/S = unuseable.

The drivers found after rebooting to the smaller partition were because you now have a 'standard PC' not an ACPI 86x-based PC as a result of changing the BIOS setting.

Good luck.