WD Data Lifeguard hangs during bad sectors repair

[hd_blues]

My trusted XP Pro desktop won't boot all of a sudden. Got message that there is a disk consistency problem with the OS hdd, chkdsk would run through 3 stages without error, and then hangs.

Installed the hdd on another XP Pro machine with OS on a different hdd. Got the same disk consistency message and chkdsk run on the problem hdd with every boot. Was able to read the data files from the problem hdd, and have backed them up. Now attempting to fix it to restore the OS and programs.

Downloaded and ran WD Data Lifeguard on the problem hdd. Quick test passed, and extended test reported bad sectors that may be fixable. Started repairing bad sectors, and part way through it hangs without any error message.

Questions:

1. What does it mean when repairing bad sectors hangs? Is the hdd physically damaged and cannot be repaired?

2. What can be done?

3. Is there a way to get an error report from the extended test? Here's what I got from the quick test:

Test Option: QUICK TEST
Model Number: WDC WD2000JS-60NCB1
Unit Serial Number: WD-WCANKC090386
Firmware Number: 10.02E02
Capacity: 200.05 GB
SMART Status: PASS
Test Result: PASS

Thanks in advance.

[northwind]

Can you give us a detailed raw SMART report?

[hd_blues]

Can you give us a detailed raw SMART report?

The only report I was aware of was the Quick Test report in my original post. After running the complex test, there was a message about bad sectors and I did not notice any other reports.

How can I get a detailed raw SMART report? Thanks.

[Vulcan]

@hd_blues,

How can I get a detailed raw SMART report?

Several utilities can collect this from a drive - these include HDDScan (free from this website) and smartctl from the smartmontools package (free from Sourceforge). It will be very interesting to see the SMART data, as northwind requested.

Just FYI, I saw this in the help file of a Windows version of Data Lifeguard:

WD also offers a DOS version of DLGDIAG, which provides a better repair function. It can be downloaded from WD web site and used to repair certain errors that are not correctable under Windows environment.

That comment from WD might be relevant to your "Windows hanging" problem.

[hd_blues]

@hd_blues,

How can I get a detailed raw SMART report?

Several utilities can collect this from a drive - these include HDDScan (free from this website) and smartctl from the smartmontools package (free from Sourceforge). It will be very interesting to see the SMART data, as northwind requested.

Just FYI, I saw this in the help file of a Windows version of Data Lifeguard:

WD also offers a DOS version of DLGDIAG, which provides a better repair function. It can be downloaded from WD web site and used to repair certain errors that are not correctable under Windows environment.

That comment from WD might be relevant to your "Windows hanging" problem.

I downloaded and ran HDDScan. Clicking the SMART button produced a report with a few yellow warnings, Reallocation Sector Count, Reallocation Event Count, Current Pending Sector Count, Uncorrectable Sector Count, and UltraDMA CRC Error Count. I have no idea what these and the associated numbers mean. How can I attach the report here?

Then I ran the Short Test from the Tasks button's menu option SMART Offline Tests. After the test finished, there was again no report. Am I running this correctly?

Thanks again.

[STeALtH]

Use the 'Upload Attachment' feature at the bottom of the reply box to upload a file.

S

[Vulcan]

I downloaded and ran HDDScan. Clicking the SMART button produced a report with a few yellow warnings, Reallocation Sector Count, Reallocation Event Count, Current Pending Sector Count, Uncorrectable Sector Count, and UltraDMA CRC Error Count. I have no idea what these and the associated numbers mean. How can I attach the report here?

Don't worry, no-one is expecting you to know what the numbers mean. If you are interested, there are articles on the web (inc Wikipedia) which discuss the various SMART attributes. We're hoping to see how bad the drive thinks its own health is, from those numbers.

Depending on the version of HDDScan you're using, then either a file called SMART.txt is created automatically in the same directory as the executable, when you click on the SMART button (so you could attach that file); or you have an option to "save to file" on the same window where the SMART attribute values were seen on newer versions (that creates a .mht file). Or if all else fails, you can take a screenshot using the standard "print screen" Windows functionality, paste that into Windows paint, save that as a jpg and attach that. In all cases, you use the "upload attachment" section, below the preview / submit / save draft buttons.

Then I ran the Short Test from the Tasks button's menu option SMART Offline Tests.

At this stage, no-one is asking you to run tests - northwind just asked for the full SMART attributes, inc raw values.

Edit: "Snap" STeALtH

Last edited by Vulcan on March 8th, 2012, 19:28, edited 1 time in total.

[STeALtH]

Hehe

[hd_blues]

HDDScan 3.1.

I saved the report file from clicking the SMART button as a mht file. But this forum won't attach it. So I opened the file and save it as a txt file, which is attached. On the mht file, NUM 005, and 196 through 199 all have the yellow warning triangle. Hope that's what you are looking for.

Really appreciate your patience with this newbie

Attachments

WDC WD2000JS-60NCB1-WD-WCANKC090386-SMART.txt

(1.12 KiB) Downloaded 1636 times

[Vulcan]

Thanks. The file was a bit mangled (lost some spaces), but I think it's showing this (slightly edited, as I don't have all night to make the columns line up ):

Code:

Num Attribute Name            Value Worst Raw(hex)       Threshold
005 Reallocation Sector Count 148   148   00000000-019D  140
196 Reallocation Event Count  170   170   00000000-001E  000
197 Current Pending Sector    200   200   00000000-0004  000
198 Uncorrectable Sector      200   200   00000000-0004  000
199 UltraDMA CRC Error        200   200   00000000-0001  000


So based on typical interpretation of attribute 5, there have been 0x19d = 413 (decimal) sectors reallocated. The "cooked" attribute value is down to 148, so it is only a little above the trigger threshold for a SMART failure, which for that disk is 140.

There are 4 more currently unreadable sectors which may be reallocated in future.

IMHO, it's a sick disk, and is close to telling you that itself via a SMART warning, when the cooked value of attribute 5 drops from 148 to 140, after some more sectors are reallocated. At least you were lucky and managed to get the data you needed from it

Personally I wouldn't trust it for anything more than scratch data. It won't get better on its own. Assuming the disk is out of warranty, and if you want to take the risk of continuing to use it, I would do what is commonly called a "zerofill" (I think WD DLG has that option), to see if the disk can reallocate the current unreadable sectors. If the zerofill succeeds, then see if the drive will pass an extended SMART test. Lots more I could try to explain, but I think it'll be a waste of time for a drive that sick, which is likely to deteriorate further. Others may disagree

Use this as a lesson for the future - you may not get this warning next time before a disk fails catastrophically, so use backups!

Anyway, I was originally just helping to answer how you could get the SMART data requested by northwind. Although my day job involves working with lots of disks, I don't work in DR, unlike him - so listen to him if he replies and says something different to me

[hd_blues]

Thanks for the comments and suggestions. I took some time to read wikipedia's explanation of the SMART attributes, and found it very helpful.

http://en.wikipedia.org/wiki/S.M.A.R.T.

I plan to use SMART to monitor my drives in the future, and have a few more questions.

- What are the meanings of Value and Worst in the report?

- What did you mean by "cooked" value?

- With some attributes such as 001 and 200, the Value and Worst counts are higher than the thresholds, and the Raw(hex) counts are zero. What do they mean?

- Are the SMART error logs permanent? I.e., they will remain there even if I completely reformat a drive?

- On a brand new hdd, should there be ANY bad sectors or reallocation?

[Vulcan]

Thanks for the comments and suggestions. I took some time to read wikipedia's explanation of the SMART attributes, and found it very helpful.

I've got limited time for replies, so I hope these help, but I can't sensibly explain everything via web forum.

- What are the meanings of Value and Worst in the report?

Those are the current value, and the lowest (i.e. worst) value which that attribute has ever been. That is useful for some attributes which can reduce & increase again. For those where "Value" will only decrease (like Reallocated Sector Count), it's usual for "Value" and "Worst" to be identical.

- What did you mean by "cooked" value?

A "cooked" value is anything which is not the "Raw" value - so those in the Value, Worst & Threshold columns are what I (and my colleagues) refer to as "cooked", to distinguish them from "Raw". These "cooked" values are scaled to fit (usually) between values of 100 and 1 (0 is not a valid value) which shows scale, but not the underlying detail. This design is explained in the history of the SMART functionality.

While not an industry standard term, "cooked" is used in some SMART discussions, to distinguish Raw from other values, as I explained. It is also used in some OS discussions, for different device driver functionality, depending on whether data is being modified or not - which also fits with the difference in meaning here. Remember that the SMART "Value" of 1 for something like Reallocated Sector Count, tells you something very different from a "Raw" count of 1 for the same attribute. So somehow people have to use terminolgy to distinguish exactly which they mean.

- With some attributes such as 001 and 200, the Value and Worst counts are higher than the thresholds, and the Raw(hex) counts are zero. What do they mean?

I don't understand - a Value above the threshold is (theoretically) "good" (at least it means that a warning is not being reported, which is not quite the same as good, but the basic SMART warning system is binary i.e. pass/fail).

- Are the SMART error logs permanent? I.e., they will remain there even if I completely reformat a drive?

Yes, they remain if you do an OS format. No, they are not necessarily permanent e.g. a firmware update could (but doesn't have to) remove them.

- On a brand new hdd, should there be ANY bad sectors or reallocation?

Interesting question. There are always imperfections which are remapped / reallocated during factory self-scan etc. but those are not usually known by users, so I assume you don't mean those. For reallocations which you can see through SMART, then I have a view. My experience talking to some disk manufacturers is that they have a different view, especially if you try to get a warranty replacement on the basis of SMART data. That's a whole different discussion...

Have fun

[hd_blues]

It's been very educational, although I still have many questions. For now that will do, and thanks for your time and patience.