Bad Blocks issue ???

I posted this question up at Western Digital for some help. I am hoping someone here might be able to assist me as well.

------------------------------------------------------

I am experiencing a problem on one of our servers. The event viewer logs reported two(2)

Errors:

Event Type: Error
Event Source: Disk
Event Category: None
Event ID: 7
Description:
The device, \Device\Harddisk0\DR0, has a bad block.

Event Type: Warning
Event Source: dmio
Event Category: None
Event ID: 25
Description:
dmio: Fail to reassign bad block(s) on disk Harddisk0: error 0xc0000010


The Server OS is Windows 2000 Server.
The Harddrives are configured for software raid : Mirror
Each drive has three(3) Volumes, C:, G:, H:
Volume 1 for both drive is C:\ (system) and is Healthy
The Windows Disk Management Console reports that drive 1 (DR0) is online (Errors)
The second and third volume on DR0 is reported as Healthy (at Risk).
Likewise, Volume 2/3 on DR1 is reported as Healthy (at Risk), although the Disk

Management console for DR1 does not indicate any errors.

I downloaded from WDC the Windows Data Lifeguard Diagnostic Utility to troubleshoot the problem. After installing and running this utility, the results indicated there was nothing wrong with either harddrive. The test results for each drive are as follows:

DR0
Test Option:EXTENDED TEST
Model Number:WDC WD800JB-00ETA0
Serial Number:WD-WMAHL1837916
Firmware Number:770.W777
Drive Type:IDE
Capacity:80.03 GB
SMART StatusASS
Test ResultASS
Test Time:08:42:35, November 22, 2004

DR1
Test Option:EXTENDED TEST
Model Number:WDC WD800JB-00FMA0
Serial Number:WD-WMAJ91102193
Firmware Number:310.G331
Drive Type:IDE
Capacity:80.03 GB
SMART StatusASS
Test ResultASS
Test Time:07:46:58, November 22, 2004

Windows is reporting a problem.
Your Utility software reports no problem.
Why is there a discrepancy and what do I need to do ?

Thanks,

Don Munyak

 

Hi Don,

I don't know Win2000 and I don't know RAID. Anyway...

Just some things you could try.
Run the Windows disk scanning utilities in Safe Mode.
Boot to the Windows disk and try disk diagnostics there.

Could try Norton Utilities Disk Doctor (read on how it may react to the RAID configuration first).

The "mirror" may not be in sync. So Windows reports an error.

Read if the Life Guard test is including a full surface scan. You may need to run another type of test. Life Guard may not actually repair the bad blocks, it may left up to Windows to do the repairs.

See if there are diagnostic utilities for the RAID. They may be able to pick up the decrepency.

Hope there is some food for thought here.
Matt

 

You might find info in this post (link below) that i answered about the same problem recently. Save me typing it al again

http://www.windowsbbs.com/showthread.php?t=37525

By the way

Bad Sectors = Bad Blocks = Bad Clusters (they are all the same thing)

Hope it helps!

 

I've not gotten much help from WD. Basically they said

Of course this was said in a very politically correct and plesant manner.

I can see this being a weekend at the office

Mattman...Thanks,your info will require a weekend in the office.
Dez...Thanks, unfortunately the advice didn't end in a solution.

I will post back my results when completed. I can already see having to break the mirror and possibly modifying the boot.ini file...argh !

 

The thread i referred you too, had many posts and a few by me, but most important was the advice about the hard disk with bad blocks. It wasnt a solution but advice to be careful if you have critical data on any drive with bad block/clusters/sectors as from experience drives with this problem often get worse not better. Just be careful and backup a lot more until you replace the drive.

Diagnostic program that repair these bad blocks only mark them as bad so the system wont write to them in the future, but dont physically fix the problem, and bad sectors is a physical fault 99% of the time that gets worse, sometimes quickly.

Just back up religiously if you dont replace it ok.....

Its a tip not a solution.

 

Thanks

Backing up is already done. This is a production file server with backup's done daily.

Once I break the mirror, I will do further diagnostic test and eventually replace the drive anyway.

The point of posting this was more for understanding. I am not, by nature a parts changer. If something goes wrong, I will typically bang at the problem until I understand it...as opposed to simply changing the part. Maybe not the most cost-effective approach, but once I have figured out the problem, it rarely comes back to bite me.

I know too many "part-changers" that will sooner blow-away an OS and start over, before understanding why a given issue occurred.

Thanks again.