HDDSuperClone

maximus

Member
Well, one I was hoping for is a 160GB image. That will most certainly be corrupt in this case. I have some other large multi GB backup recovery files that I am not even sure would be useful now even if they were not corrupt, so I am not even going to bother with them. And there are some other test verification files of 4GB. So the file size I am talking about is 4+ GB, and I consider them useless to me if they are in any way corrupt. But none are critical data, just things I would have liked to keep at the moment.
 

maximus

Member
jol":2plber79 said:
[post]4491[/post] what FS is it ?
NTFS
And it doesn't matter any more. I have coped what I think I wanted to keep, and have now reformatted the drive that it was cloned to. I can't tie up the good drive any longer so the deed is done.
 

maximus

Member
From this lesson I have added an extra copy phase. Phase 1 and 2 are still with the head skipping algorithm, but now phase 3 is a forward pass with fixed skipping based on the read rate. It is the last chance to get more data from a weak head fast. Phase 4 is now what phase 3 was, which is without any skipping.

And I have also found the hardest, if not totally impossible thing so far. I am attempting to produce an estimated time remaining for recovery. There are so many variables that it can never be accurate if even anywhere near close, but I am attempting to at least make it a ballpark figure. The hope is that it will be able to give a basic idea of whether it will take hours, days, weeks, or longer.
 

maximus

Member
Don't get too excited as this is just the free version, but I have released an alpha testing version. But I do need some specific testing of this. I need it tested on a few drives that will go into device fault (disappear from the likes of ddrescue). I need to know if it will exit with an exit code of 127 (0x7f) when the drive goes south on reading bad sectors. The exit code is displayed on exit of the program, plus it is supposed to give a message if a device fault was detected.

http://www.sdcomputingservice.com/hddsuperclone/alpha
 

maximus

Member
I made a few modifications already, so if anyone downloaded the initial 0.1-alpha release, you should download the newer 0.2-alpha release. It now will also exit in other circumstances where the drive goes bye-bye with an exit code of 63 (0x3f). Version 0.1 might not exit and continue with the operation while reading no data. I tested it against ddrescue by dropping the power on a SATA connected drive, and also removing a USB drive and both would exit. But I found that if I pulled the drive off my WD Passport board and left the USB plugged in, ddrescue would hang indefinitely until the USB is removed and then exit, while my program will exit with the error code every time. Have yet to test it on an actual IDE drive, but expect the same results as SATA.

Oops, just realized that it won't mark the chunk as bad in the event of a device fault, that will be fixed in 0.3. But I still need to know if it will exit like it is supposed to on a device fault.
 

maximus

Member
Hmm, I was wrong about IDE results being the same. I have made it so that it will recheck the drive size after every error, which seems to come in very handy for passthrough mode. When the power is pulled from an IDE drive, hddsuperclone will exit. Ddrescue however, will continue on while marking all data as bad. :eek:

Edit:
I found that ddrescue would even restart again on the IDE drive that had the power removed, and continue to mark all as bad. Even an fdisk -l command still showed the drive with the original reported drive size. But Hddsuperclone refused to start cloning as it reported the drive size as 0. :)
 

Jared

Administrator
Staff member
I've seen a lot of IDE drives that take up to two minutes to ID, but after that they work just fine. I guess they are just from a time when computers booted up much more slowly giving you time to get a cup of coffee and possibly drink it before going to work. :D
 
Top