FBLite - ABA and checksum errors in SA. Cannot correct

[avalanche]

Hello all,

I wondered if there were any suggestions on how best to approach this HDD. It is proving challenging. As received the HDD would spin up, stay spinning, reach RDY, and provide the model number but no capacity. Mod 11 would not load from the platter surface, but 01 Dir would. After loading Dir and donor 11 to RAM it was possible to access SA. Attempts to read the SA showed lots of modules reporting ABA errors. At this stage, I thought it best to check the heads. Donor heads were used and the fault it replicated. The patient heads work fine in the donor too, so we can rule out a heads issue.

Fortunately, all of the "A" and "As" modules read OK, but there are errors in a number of the "B" modules. For copy 0 it has been possible to write donor modules successfully to the patient and verify these OK. However, for copy 0 when I write mod 0023 and 0024 (Smart Log reserved) they fail the checksum test. If I do the same for head 1 the checksum is passed and the module is reported OK. Likewise if i try to correct certain "B" modules on copy 1 many are OK, but some fail checksums. In short I cannot get either copy 0 or copy 1 to be completely OK for the "B" level modules.

Next I considered that all of the previous ABA errors and slow reading of the SA might be due to degraded areas of the SA, and that editing the DIR to re-assign modules 0023 and 0024 to a different ABA range would be useful. However, when I make the appropriate changes to DIR and write the edited 01 to copy 0 and copy 1, I then get a checksum error on copy 0, but OK on copy 1. If I use the CS Recalc option it corrects the module. However, when the module is written back to the HDD it reports success, but then fails the verification, again reporting a checksum error.

I have not attempted writing the patient modules to a donor HDD in preparation of a hotswap yet. I thought I would leave that as a last option if it is not possible to repair the SA on the patient HDD.

I've attached the log for module check on copy 0 below before any changes to the SA were made.

Thanks for any suggestions,
John.

------

[GuideRecovery]

by the description it is clear the Media is degraded so smart Hot swap is the only option i think.

[AJ2008]

Source of the post by the description it is clear the Media is degraded so smart Hot swap is the only option i think.

you will have issues to hotswap this due to extra adaptives these models use.

Best option here is to find some 'unused' areas in SA, or to shrink some non important module and allocate its space to some modules you cannot repair and relocate these modules.
Depending on which modules are affected and how much of the module is readable you may be able to simply shrink some modules and ammend checksum and this could be ok.


So - read all modules in composite
identify modules that you need for data access
use DIR editor to shrink or relocate modules to other area
for some modules you cannot use from donor, for others you can - you need to work this out.

[avalanche]

Composite read of modules took place overnight. It did not improve results massively, there are still lots of modules with error.

Yes, I'm keen to avoid attempting a hotswap on this series of HDD.

The issue I am facing is that when I alter DIR 01 to reallocate certain modules to unused areas of the SA, is when I write the new 01 back to Copy 0 and Copy 1 of the SA there is then a checksum error on copy 0, but not copy 1 of the written 01 dir module. Consequently I cannot save the changes made to dir to alter the ABA locations of modules. Frustrating!

John