HDD GURU FORUMS

maximus wrote:

The image shouldn't grow because the tool should already had created the image to the full size with zeros or "bad" or whatever pattern filling the space on the image that could not be read on the source drive.

Technically it uses sparse writes. If you write the last sector to the file Linux will show it as full size. But if you use a command that shows actual disk usage, you will see that it is only taking up the space that has been actually written.

Code:

du -ah

I do have a suggestion for your Cloning tool ! I don't know if you want to implement something like this but it would be nice to have, at least on the paid version.

When you skip bad sectors you might want to reverse clone/image from the point where the skip ends backwards to the first bad block it enconters.

Example - You are cloning a drive and you hit several bad sectors so you skip a pre-defined number of LBAs. If you end up on a sector that is good then instead of moving on to the next LBA you should go back and try to clone/image the prior LBA and so on untill you reach a bad sector. Then you would move forward again. This might help with your bad head calculation algorith as well. let's say that you are imaging from LBA 0 to 1000 and you set a skip of 100 sectors when 5 sectors are bad in a row. If you have LBA 10 to 20 "bad" then your tool will reach LBA 15 and will skip to 115 (or 116 or whatever). then instead of moving on to 117, 118, etc .... it would instead attempt a reverse clone from 115 backwards untill it reach the 20 (that would be the first bad block in reverse from the point you skip). This would maximize the data that you would grab from a big "skip". This is the same that HRT-DRE does (unless you turn it off). If you can implement something like this and if you end up cloning a drive with a bad head it might help you out to skip the head and re-adjust by going back to get the data if the skip ends up on a further point of the working head instead on the first LBA assigned to the working heads.

Spildit wrote:I do have a suggestion for your Cloning tool ! I don't know if you want to implement something like this but it would be nice to have, at least on the paid version.

When you skip bad sectors you might want to reverse clone/image from the point where the skip ends backwards to the first bad block it enconters.

Example - You are cloning a drive and you hit several bad sectors so you skip a pre-defined number of LBAs. If you end up on a sector that is good then instead of moving on to the next LBA you should go back and try to clone/image the prior LBA and so on untill you reach a bad sector. Then you would move forward again. This might help with your bad head calculation algorith as well. let's say that you are imaging from LBA 0 to 1000 and you set a skip of 100 sectors when 5 sectors are bad in a row. If you have LBA 10 to 20 "bad" then your tool will reach LBA 15 and will skip to 115 (or 116 or whatever). then instead of moving on to 117, 118, etc .... it would instead attempt a reverse clone from 115 backwards untill it reach the 20 (that would be the first bad block in reverse from the point you skip). This would maximize the data that you would grab from a big "skip". This is the same that HRT-DRE does (unless you turn it off). If you can implement something like this and if you end up cloning a drive with a bad head it might help you out to skip the head and re-adjust by going back to get the data if the skip ends up on a further point of the working head instead on the first LBA assigned to the working heads.

That is what phase2 does, it copies in reverse with skipping to get any good data that was missed due to being skipped in the forward pass. Also, the self adjusting skipping algorithm does its best not to over skip by too much, meaning it only skips as far as it thinks it needs to. You should try it sometime on a drive with a weak head and watch the pattern it makes (use hddscviewer for the visual reference).

maximus wrote:(...)That is what phase2 does, it copies in reverse with skipping to get any good data that was missed due to being skipped in the forward pass. Also, the self adjusting skipping algorithm does its best not to over skip by too much, meaning it only skips as far as it thinks it needs to. You should try it sometime on a drive with a weak head and watch the pattern it makes (use hddscviewer for the visual reference).

OK !!!
Many thanks !

Hi,

so many days but only grow 1gb, see

20 74 68 65 73 65 20 64 65 66 61 75 6c 74 73 2c these defaults,
20 74 68 69 73 20 64 72 69 76 65 72 20 77 69 6c this driver wil
6c 20 6d 61 6b 65 20 31 20 68 6f 73 74 20 77 69 l make 1 host wi
Source: /dev/sdb Destination: /mnt/2/root/intento2/160gb.img
Total LBA: 312581808 LBA to read: 312581808
Run time: 10:09:03:22 Remaining: 191:14:38:08
Rate: 1174 B/s Recent: 4888 B/s Total: 2046 B/s
Skip size: 4096 Skips: 35891 Runs: 0 Resets: 0 Recent: 0
Position: 6544102 Status: Phase4
Finished: 160632424 (75903 areas 51.388924%)
Non-tried: 149792266 (50458 areas 47.920979%)
Non-trimmed: 2096607 (15625 areas 0.670739%)
Non-divided: 0 (0 areas 0.000000%)
Non-scraped: 0 (0 areas 0.000000%)
Bad: 60511 (48252 areas 0.019358%)

I think that you should stop it.

Take the drive to a data recovery firm if the data is important to you.

You have done all that you could with free tools.

I can see only 2 options.

Either the drive clones "better" using hardware assisted tools or the drive needs head stack replacement.

Just to rule out for sure PCB problems like dying MCU did you change the PCB of the drive + ROM ?

enito wrote:Run time: 10:09:03:22 Remaining: 191:14:38:08

It is on track to take a good 6 months to finish, probably longer. That is a long time, if the drive were to even last that long. It is difficult to tell if it would image much better with a hardware imager without trying, but my money would be on requiring replacing the heads, not a DIY job.