WD Zeus and Hulk. Healthy SA - but no access to surface

[ACE_Admin]

I met several cases with Hulk and Zeus WD drive familes - where all SA Structure test had no errors, but the drive did not allow to get the data from it.

We do not have a ready solution for this problem, but our developers are working on it. If you have a similar problem with these or other WD Marvell families, please provide us the SA from them. That will help the developers to find out this problem's solution.

Thanks in advance.

>>AJ2008

Please make backup of your patient's SA(modules and tracks) and attach to your message or put somewhere on a filehosting and provide a link.

[AJ2008]

I have had 2 of these in 3 weeks

[AJ2008]

Also, this may not be important, but all my HULK drives with this problem have lots of difficulty and ABRT errors when reading Service tracls from Cyl -1, whereas all donors do not have this trouble.

[ZeBong]

i Have one case off the Tornado 3D family , dont support yet by the pc3k.

It detetct the ID but could not copy any sector.

When i test the SA it was ok, when i test the heads it was Ok.

Also i post some info about this drive.

ZeBong

[AJ2008]

Seems like this fault is becoming quite common. Maybe it is going to replace the 7200.11 failures!

Theres not much talk about this fault at the hddguru forums, I would have expected something there!

[latch]

Similar fault with a Tornado (WD3200AAJS). No faults found in SA and heads test completed ok with no errors.

In Data Extractor, I could reverse image the drive ok and successfully imaged approx 70% of the drive, but first 30% was generating errors on every sector and direct sector access was causing the drive to hardware reset and re-initialize.

Modified DE config to set "Block size for reading" from 256 to 2 or 4 and it is now imaging ok, albeit slowly.

[ZeBong]

Modified DE config to set "Block size for reading" from 256 to 2 or 4 and it is now imaging ok, albeit slowly.

Good to know that

[Monk]

Seems that Kermit drives (WD10EACS) also have this problem. Anyone tried a hotswap ? don't have good drive at the moment to try this.

[AJ2008]

It seems as though this fault is related to one or more heads being faulty causing translator to not load completely giving limited or no access. I have done HotSwap on Kermit with a similar fault, but never managed success on Zeus or Hulk yet.

[dobrevjetser]

Same problem with Jupiter family; i've got two cases in; SA ok, but no data-access.
All heads tested OK.
But indeed on both Abrt on the sectors 263 - 1301 from SA.

Dobre

[Balvan]

I've fixed this problem on Tornado & Tornado 2R by replacing PCB.

In several cases, I could read SA but not client data. Swapping PCB fixed it.

I would like to know the reason for this. One might think that if you can read SA, read channel, etc. is working just fine.

Balvan

[GURU]

I have a feeling that the SA area on these S4W drives are hard-coded (Servo written old style) And the user area is servo written in auto self write mode.. Maybe harder to lock on to this area if adaptives are incorrect or if problems arise on surface or with head quality.

Does the G-List show a high amount of defects on certain heads and zones?


AND a silly question does the drive ID show unlocked state?

[AJ2008]

I had a 1TB HULK with this fault. PCB replacement made no difference, GLIST showed no unusual amount of defects, and all heads passed head test, but on firther inspection no one part of the disk could be read using head 7, this head is damaged and because of this fault the translator could not completely load causing ABRT on accessing.

Head Replacement or HotSwap, both very difficult, is only solution.

[dobrevjetser]

Anyone got allready a solution for this ?
I've got two WD7500AAKS with this problem; one after a headswap, and one after replacing the pcb (smooth was burned and processor dead) but repaired with modules from the SA.
Both drives are correctly identified and SA can be read without problem, but no access to data whatsoever.
Translator regenerating is not helping.
Any help appreciated.

Dobre

[sarmich15]

Was anyone able to fix this problem? We have got a couple of 1TB Hulks with the exact same symptoms...
Thanks a lot!