ST31000640SS powering off on bad blocks

[florian.bobelka]

Hi Guys!

I hope someone here can help me with a raid-dissaster...

The Story:
A customer has a RAID5 Array consisting of 4 HDDs of type ST31000640SS in a Promise VessRAID System. Yesterday one drive (the 3rd/"SeqNo2") failed. The global spare drive (a SATA HDD) immediately jumped in and AutoRebuild was startet to synchronise that new drive. While this task was running at ~6%, a second HDD (the 1st/"SeqNo0") also failed, AutoRebuild stopped and the whole array with all volumes went offline.

I tried to reset "SeqNo0" by reinserting and forcing online - this worked so far.
Then I started the synchronisation to the spare drive manually - at ~6% "SeqNo0" failed again.
Reading the event logs of the VessRAID showed that there were some warnings regarding bad blocks on that drive immediately before it failed.

I decided to connect "SeqNo0", "SeqNo1" and "SeqNo3" - the HDDs with the most accurate Data - to a SAS host controller in my technician computer and working with Runtimes RaidReconstructor and GetDataBack as I have very good experience with those tools.

Unfortunately, every time the readout of GetDataBack reaches the bad blocks on "SeqNo0", the drive powers off. Repowering the HDD brings it back online, but still running GetDataBack can't find/access it after that. Closing and restarting GetDataBack gives again access to the drive, until the BadBlocks are reached and the drive powers off again...

What I would now like to do is to take a bytelevel image of the bad HDD (to not torture it more that it is needed). Bad blocks can be ignored, but imaging MUST run to the end of the drive.

Do you have any idea what I can do?
Is there a tool to tell the HDD to NOT power down?
Or a imaging tool which pauses and finds the drive after repowering?
Even a payable data recovery service in/near Austria would be of great help!

I hope you have any ideas,
thanks for reading,
Florian Bobelka.

[northwind]

Frankly, i think that powering down the hdd due to bad blocks is very unlikely.
If i were in your shoes, i would try to take a sector-by-sector image of the drives using Disk Recoup and then analyze the images in order to get data.
If this fails, then a skilled pro should have a look at the drive in question, in order to have it cloned.

members hddguy and dr-kiev of this forum can assist you in analyzing the raid and they come highly recommended.
Good luck!