WDC WD2502ABYS-18B7A0, FW 02.03B05

[Gary]

Background:

I got 5 drives in from a 4-drive RAID-5:

1 = Loose drive: Failed drive that was in Slot1 that was replaced in January

2 = Slot0: Working drive

3 = Slot1: Blank drive (all sectors contain 0's)

4 = Slot2: Working drive

5 = Slot3: Recently failed drive (this is the drive I am working on)

The IT guy for the client replaced the drive from Slot1 in January because it had failed. Since they are hot-swap drives, he thought all he had to do was swap drives and it would rebuild itself, so no rebuild was initiated.  Now the disk in Slot3 failed. I confirmed that Slot1 was blank and that Slot3 was failed (clicking). Drives from Slot0 and Slot2 imaged fine. When performing data recovery on RAID sets, you usually have donor drives  that come with the recovery since they are usually same make, model, firmware even exact same date, etc. But in this case 3 drives are Dell branded Seagate drives and 1 drive is a Dell branded WD. Of course the drive that I need to fix is the WD.  It is a less common WD2502ABYS-18B7A0 WD RE3 drive. That is the background, now on the the issue.

When the drive came in, it was clicking and would never become ready. After troubleshooting I determined head issue. I was able to find a pretty close donor drive. Exact same model, but firmware was slightly off (02.03B05 FW on failing drive and 02.03B04 FW on donor drive). Donor drive came in and tested fine, so I used PC3000 to backup resources on donor drive. I swapped heads. After powering on the drive it immediately spun up and became ready. So I backed up resources on this drive which succeeded. I did get a header error on module 00, but I got the same error on module 00 on the donor drive which worked fine so I didn't worry yet. I put the drive on the DDI and all sectors were timing out. So I cancelled and jumped to pass 2 which gave all ABORT errors. I cancelled again and started a pass reading-ignoring-ECC. This was working (DDI giving the green ECC status for each sector) but all sectors looked like they contain about the same data. I put the drive back on the PC3000 and tried to read sectors and it gives me ABORT for each sector I have tried.

I think the head swap went fine since the drive went from dead to being recognized (model, SN, size all OK) and PC3000 can back up all modules except 00 which returns a header error, all service tracks, etc. Does anybody know what module 00 is or what the issue might be?

[Gary]

Update: After playing with the PC3000 I have it reading data. I think that head 1 was too weak to properly initialize during the drive startup, so I edited the heads map in RAM (changed to 0 0, soft reset and then change back to 0 1) and the drive is now properly reading the user data area. After the drive finishes imaging I'll confirm that this worked.

[Gary]

Editing the heads map in RAM worked. I was able to finish the image of the drive with 55 sectors unread (UNC that couldn't be read). Here is what I did in case it helps anybody else:

Since I don't have the PC3000 Data Extractor, I had to "hot swap" the drive to the DDI.  To do that, I moved the 2 machines closer together (so the SATA cable for the DDI could reach the drive while hooked up to the PC3000). I powered up the drive in PC3k, verified it couldn't read sector 0, 1, 2, etc. and edited the heads map in RAM as listed above and then verified it could now read sectors 0, 1, 2, etc. Then I kept the drive powered on through PC3000, but swapped SATA cables so the drive was still powered by PC3000 (can't power cycle or RAM changes would be lost) but data cable was now on DDI. DDI saw drive properly and was now able to image properly. I did heads map in DDI and first imaged head 0 which went very fast (less than an hour for 250 GB drive) without any errors. Then I had it image head 1, which got slow in a few places, but completed in about 3 hours with only 55 unread sectors. I then imaged the 250 GB of sectors on the DDI destination disk and used that with the other 2 good disk images to rebuild the virtual RAID-5 (I used EnCase, but could have also used R-Studio, RAID reconstructor, etc). It was able to see the 3 partitions and all data appears valid.

[Alandata Recovery]

I use the same hot swap trick.

I have a long sata cable and use the power from the ddi to power the drive.

I use the computers power supply - to prevent ddi from repowering it.

Another trick is to use a small power supply for the sata power - the kind used for powering 1 drive that comes with the usb adapters works fine.

I plug it into a small ups then power up the drive and get it ready on the pc3k.

Now I can unplug it and walk the whole thing into another room.

On Thu, Jul 2, 2015 at 8:02 AM, Gary <ghue...@houstoncomputing.com> wrote:

Editing the heads map in RAM worked. I was able to finish the image of the drive with 55 sectors unread (UNC that couldn't be read). Here is what I did in case it helps anybody else:

Since I don't have the PC3000 Data Extractor, I had to "hot swap" the drive to the DDI.  To do that, I moved the 2 machines closer together (so the SATA cable for the DDI could reach the drive while hooked up to the PC3000). I powered up the drive in PC3k, verified it couldn't read sector 0, 1, 2, etc. and edited the heads map in RAM as listed above and then verified it could now read sectors 0, 1, 2, etc. Then I kept the drive powered on through PC3000, but swapped SATA cables so the drive was still powered by PC3000 (can't power cycle or RAM changes would be lost) but data cable was now on DDI. DDI saw drive properly and was now able to image properly. I did heads map in DDI and first imaged head 0 which went very fast (less than an hour for 250 GB drive) without any errors. Then I had it image head 1, which got slow in a few places, but completed in about 3 hours with only 55 unread sectors. I then imaged the 250 GB of sectors on the DDI destination disk and used that with the other 2 good disk images to rebuild the virtual RAID-5 (I used EnCase, but could have also used R-Studio, RAID reconstructor, etc). It was able to see the 3 partitions and all data appears valid.

--
You received this message because you are subscribed to the Google Groups "DataRecoveryCertification" group.
To unsubscribe from this group and stop receiving emails from it, send an email to datarecoverycertif...@googlegroups.com.
To post to this group, send email to datarecovery...@googlegroups.com.
Visit this group at http://groups.google.com/group/datarecoverycertification.
For more options, visit https://groups.google.com/d/optout.

--

Alandata Data Recovery -  (949)287-3282  
"Cleanroom Data Recovery of RAID, VMware, NAS, Linux, Tape, Disk, Forensics"

www.alandata.com   www.alandataRecovery.com

[Philip Shaw]

That sounds pretty cool but wouldn't have been easier to get Data Extractor? I use it all the time and it offers a whole lot of ways to get the customer's data back.

[wayne horner]

isnt data extractor $2500 ?

I guess thats easier.

Alandata Data Recovery -  (949)287-3282  

"Cleanroom Data Recovery of RAID, VMware, Network Attached Storage, Linux, Tape, Disk, Forensics"

www.AlanData.com   www.AlanDataRecovery.com

[Gary]

Yeah, Data Extractor would be easier. I still need to get it.  And the SSD add-on.  And the Flash add-on. And the PC-3000 for SAS, etc.