Markers Table

[ddrecovery]

I have been using VNR for a while, but have to admit I have not really concentrated on it enough to get full use out of it. As we have a support contract with Rusolut its easy to get the guys to log in, but I don't learn much that way. So I am about to post quite a few questions here in the hopes that I (and maybe a few others) can learn a little from real examples.

I have a job now that has the markers table as the attached picture. There are missing markers and duplicated markers. I can get some folder structure but it is pretty corrupt but at least I know I am on the right track. This case has the PAIR function included. On the VNR class most of the marker tables are pretty easy to do, ie start at 0000 through to XXXX with the odd duplication. But its when I get these mixed marker tables that I have issues. If anyone is wiling to help I am sure a lot of us could learn.

Thanks in advance.

Attachments

[HaQue]

Duplicated markers can be because of different banks,

(making up numbers, so YMMV)
for example, you may have one set of markers in an offset range of 0x0 - 0x100000000, then the duplicated ones from there upward... being bank 0, and bank 1

[ddrecovery]

Duplicated markers can be because of different banks,

(making up numbers, so YMMV)
for example, you may have one set of markers in an offset range of 0x0 - 0x100000000, then the duplicated ones from there upward... being bank 0, and bank 1

I could understand that if the whole table had duplicate markers. But these duplications are sporadic. I know how to remove the duplications, but not sure how to select which one to remove. Also most of the markers miss an alternate number altogether in the sequence, this happens pretty much though the who table. Just trying to understand why? ie.

00DA
00DC
00DE
00E0

[HaQue]

This could be what other software calls "conflicts" you could possibly remove one, check if results are better, if not remove the other and compare. Is this marker the Block Number? There could be another marker for version... a marker that denotes that this is the current version of the block. There is also controllers that may not actually have any way to differentiate between blocks with same markers using the dump alone... maybe a table is stored somewhere in the dump or controller itself (dump being more likely) I don't have a great understanding of this either!


some controller firmware developers seemed to have been on the juice at times.

[arvika]

What controller?
For some controllers you can filter good data blocks by other marker, here probably it will be first byte from test1 marker (1E). I see also, that LBN markers is multiplied by 2. After it if you have empty blocks in list, you need add it to avoid shifts in image. If you have same LBN markers it is called conflicts, need resolved it manually. If ECC is not good in service area, some markers can have other value (looks like lost blocks or conflicts).

[ddrecovery]

Thanks for the help. I think this is too complex a case for me to learn much from. VNR support logged in this morning to help and its still a work in progress. Gosha was changing block sizes and various other methods to improve the recovery, but its still a work in progress. I will ask you guys a similar question when I get my next one if you don't mind. Again thanks for the input.

[arvika]

What controller?

[ddrecovery]

What controller?

Sandisk 20-82-00388

[arvika]

Ah, not so easy for learning

[HaQue]

I see other software assembles this by sector number. Sorry I don't know much else about it

[pcimage]

Ah, not so easy for learning

Agree!

Sandisk can be pretty tricky, with some horrible and complex “mixes”

Not for beginners

[ddrecovery]

Not for beginners

Thanks

[pcimage]

Not for beginners

Thanks

Sorry Tim, didn’t mean to malign you! Maybe “beginners” was probably the wrong word there!

Tbh, even Sergey (Soft Center) and the guys at rusolut have problems with these. Acelab TS have never been able to solve a tricky Sandisk for me!