ST9750420AS -suspect bad head. very slow, no data access

[joshAU]

Hi all.

I have a customers ST9750420AS 2.5" hard disk that is running very slow and comes up as an un-allocated partition under disk management.
Victoria fails to read any sector, although it does seem to retrieve the SMART attributes, albeit very slowly.

Using terminal, on spin up, the drive returns the following:

EarlySpinup Enabled
Rst 0x10M
(P) SATA Reset

RAW OFF
PASS
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087
ProcessRWError -Read- at LBA 001194A5 Sense Code=40000087

at which point it stops displaying anything further.

I can then get to F3 /T by pressing Control-Z
Which returns:

APM disabled by SPort
ASCII Diag mode
F3 T>

Checking the head resistance within terminal, X, it returns:

Head 00 Resistance 0203
Head 01 Resistance FFFF
Head 02 Resistance 0306
Head 03 Resistance 02D5

Head 01 seems to be damaged, and it intermittently reports either FFFF, 0004, etc.
The other heads all have a non changing, significantly higher, resistance
Does this seem to be correct to you?

Glist, V4 returns perhaps hundreds of entries.
NRG, V40 returns:

F3 T>V40
Nonresident GList 12 entries returned
Total entries available: 12
PBA Len Flags Phy Cyl Hd PhySctr SFI
65935E3 1 20 19275 1 24 86A88
960CE4C 1 20 27769 3 76 54A13
960CE4F 1 20 27769 3 79 5700C
960CE52 2 20 27769 3 7C 5964F
960CE55 1 20 27769 3 7F 5BC92
960CE57 2 20 27769 3 81 5D5CC
960CE5A 3 20 27769 3 84 5FC0F
960CE5F 1 20 27769 3 89 63BD4
960CE61 1 20 27769 3 8B 6550E
960CE65 2 20 27769 3 8F 68812
960CE68 1 20 27769 3 92 6AE55
960CE6A 3 20 27769 3 94 6C78F
960CE77 1 20 27769 3 A1 76D14
960CE7D 1 20 27769 3 A7 7B99B
960CE82 1 20 27769 3 AC 7F918
960CE87 1 20 2776A 3 0 EAA7
960CE89 1 20 2776A 3 2 103DF
960CE8B 1 20 2776A 3 4 11D61

F3 T>

Any other suggestions as what I could be missing?
To me I'm assuming a head swap, but just wanted to double check.
Any help greatly appreciated.

[northwind]

40000087 means a reallocation has been encountered.

Assuming a visual check on the heads is not viable, I would switch off all background activities and then check heads' reading abilities.

Considering this is a client's drive I would kindly suggest you outsource this case to a pro lab, unless you have clean room facilities and relative practice. Hammering the drive can lead to non-recoverable data.

[joshAU]

Hi Northwind.
Thanks for your reply and advice.
Don't you think head 01 has failed though, given its resistance figures?

In regards to it being a customers drive, yes, with drives we cannot recover data on, we send them to Kroll Ontrack in Brisbane.

If the customer doesn't want to pay the sort of figure Ontrack charges, we usually give customers the option of risking us attempting the recovery ourselves, at a greatly reduced chance of success than Ontrack would, but also substantially cheaper.
We also always advise them that, by doing that, we could make any further attempts at recovery less likely, or even impossible.

[pepe]

i guess there is a way in the middle by outsourcing to somebody having the correct setup and skills.

pepe

[joshAU]

Thanks for your input Pepe.

[laptokowiec]

Switch off all background activities as northwind said. There's more about seagate terminal opitions on http://www.hddoracle.com.
I assume you could recover about 75% of data using ddrescue or hddsuperclone without replacing heads.

[pepe]

Head 01 Resistance FFFF

this does mean something. A partial recovery might be possible from other heads but there's a risk of killing the whole drive if surface 1 has some damage. Just keep in mind.

pepe

[joshAU]

Sorry for the delay.
Thanks for your input @laptokowiec.
We were able to get about 10% of the data.
The customer is going to collect what we've got, and then decide if they want to send the drive off to our DR partner for further recovery.
Appreciate your advice, thank you.

Thanks again @Pepe for your advice as well, appreciated.
We're going to send it off to our DR partner, if the customer decides the files are worth it.