Bad Sector Issue / TLER Issue on RAID 10 array

[Raid-Rash]

Hello,

- I have an 8-Drive RAID 10 array on a3Ware 9650SE card.
- The drives are wd5000bpkt (and so not RAID edition).
- One of the drives in the array has developed a bad sector at about the 25% mark.
- It appears that whenever the RAID adapter tries to read the sector the drive attempts to remap the bad sector but that TLER times out and drops the drive
- The mirror drive to the bad drive is in the process of being rebuilt. Each time a rebuild is attempted it hits the bad sector and apparently triggers the bad drive to be dropped resulting not only in the rebuild failing, but in the array becoming "unusable".

Clearly using non raid-edition drives was a bad decision and I need to re-think that strategy. For the moment, however, I am focusing on recovering the array.

I imagine that if I can cause the bad drive to re-map the bad sector that the array will become usable again - however I am not sure what software tool to use for this?

I have found that I cannot simply mount the bad drive in an available system and chkdsk /r because outside of the array the drive appears to be an uninitialized disk with no volumes defined.

I used the HDDRawCopy utility to copy the bad drive to another drive (a WD1002FBYS) - however the copy drive was not accepted as part of the RAID array.

Can anyone point me to a software that can resolve the bad sector issue or suggest a better alternative to resolving this issue?

RAID-RASH

[labtech]

1) Do you care more about the data or for the array to operate back as normal?
2) If care more about the data, then on a scale of 1 to 10, how important is the data on the array?

[Raid-Rash]

Result of running "Western Digital Data LifeGuard Diagnostics" Extended Test with attempt to repair sectors:

Test Option: EXTENDED TEST
Model Number: WDC WD5000BPKT-00PK4T0
...
Firmware Number: 01.01A01
Capacity: 500.11 GB
SMART Status: PASS
Test Result: FAIL
Test Error Code: 08-Error was detected while repairing bad sectors.
...

[Raid-Rash]

Do you mean all of the data or some of the data? It is difficult to place a simple numeric value on the importance of the data, but lets say it is a 5..

The value of restoring the current array lies in becoming operational again sooner.

The value of the data can be defined in terms of the time it would take to recreate the 12 or so hyper-v virtual machines stored on the array and the inconvenience of losing several weeks work (most recent backups ~2-4 weeks old).

I guess I would pick the option that requires the least amount of time for the combination of recovery and regeneration of data.

[DR-Kiev]

3ware is a good raid controller , it doesn't allow you to use any other alien drive just as is .
Moreover , it is usualy locking drives for himself only .

If you don't want to cause any others troubles, you have to clone all drives and reconstruct the array virtually, but you need to avoid 3ware lock for each drive which lead you to proper tools in any case.
The other option is to outsource this case to professionals .

Few times rebuilding process and chkdsk was wrong steps .

[Raid-Rash]

Good news!

I figured this out myself with the help of "Hiren's BootCD" and a program called "Victoria" that remapped the bad sectors for me.

Now I just need to find a better storage strategy to avoid a repeat.

Any suggestions?

Note: I originally switched to using the 8x 2.5" WD mobile drives instead of 16x 3.5" WD Raid Edition drives to try to save on my electric bill - and the savings have indeed been fairly substantial - about the cost of a new mobile drive each month.

WD did not appear to offer any SATA mobile drives that are Raid Edition (except some very expensive SAS mobile drives)..

[lcoughey]

Consider yourself lucky. That stunt could have caused the loss of all the data. As it stands right now, you likely lost the data that was contained within the remapped sectors. So, if you now have a corrupt database, you know why.