HDDSuperClone

[nissimezra]

@maximus, This seems to work like charm for wd that one of the head isn't reading.
I have a drive that if it connected to windows it mount and disconnect repeatedly.
I cloned the drive, 75% went smooth but 25% aint moving. Now it's scrapping. still only 75%.
This seems to get job like head mapping.

The smooth part would be the head skipping algorithm doing its job in phases 1 and 2. So in this case of a dead head, you basically already had the best recovery at the end of phase 2. The trick is to be able to understand the results. Someone that is technical could use hddscviewer to get a better idea of what the recovery looks like, and even tell if there seems to be any data readable in the bad head. It also helps to be able to understand the log file (by reading the manual), as analyzing it can also give an indication if any data is being read from a bad/weak head. Learn how to understand the results and you could tell at the end of phase 2 what the condition of the drive is.

And you are using the free version. Imagine using the pro version and properly implementing the timeouts for directly connected SATA drives, and making the whole process faster.

Thank you for the input. I will need some time to read and learn how to use it.
I skipped phase 4 and passed to the trimming. It is reading.
Thank you again

 

[maximus]

By popular request, there is now an available live cd based on lubuntu 18.04 64 bit. This is for those that want to access NVME SSD drives and are finding the 14.04 version is not able to recognize them.

And also here are the recent updates, mostly for the pro version.

HDDSuperClone Changelog

version 2.1.12 20190427
* Fixed a bug with AHCI where drives listed with no data
* Fixed a bug where SMART data would fail after stopping

version 2.1.11 20190423
* Fixed a bug with drives visable to the OS in direct mode
* Added a program script that can install program dependencies

version 2.1.10 20190419
* Improved SCSI write error detection
* Changed the direct SCSI write timeout to 2 seconds
* Fixed a bug with direct AHCI for some motherboards
* Updated some info in the WD VSC scripts
* Added the script wd_royl_write_mod

version 2.1.9 20190328
* Added a section in the user manual for relay operation
* Improved the SCSI read mode support detection
* Added a 1 second timeout for direct SCSI writes
* Added extra info about direct SCSI writes in the manual
* Added a translated Russian language pack

 

[nissimezra]

Wow!
Tested the 64 Bit ISO, when opening hddsuperclone there are lots of errors. Something about proxy. I'll test it today more.

 

[maximus]

Don't worry too much about the errors you see in the console when starting hddsuperclone. Some of that is from having to run it as root (sudo). You would be surprised how many console errors there can be when starting a program from the terminal in Linux. I have seen many errors from the likes of Firefox and R-Studio, but they work fine. The average user would never see that because they started it from an application shortcut that does not open a terminal.

 

[nissimezra]

Don't worry too much about the errors you see in the console when starting hddsuperclone. Some of that is from having to run it as root (sudo). You would be surprised how many console errors there can be when starting a program from the terminal in Linux. I have seen many errors from the likes of Firefox and R-Studio, but they work fine. The average user would never see that because they started it from an application shortcut that does not open a terminal.

Confirmed working with E2B.

Thx

 

[maximus]

HDDSuperClone version 2.13 released.

HDDSuperClone Changelog

version 2.1.13 20190804
* Fixed a bug with the timing byte
* Fixed a bug with some USB drives bad reads reporting as good
* Fixed a bug where SCSI passthrough would not work on old drives
* Fixed the generic source device mode
* Fixed a bug with the model name in generic source device mode
* Changed some of the default timeouts to a higher value

 

[maximus]

So I have been working on my Direct USB mode, which I now have at the beta level, and here is some information.

First, this is meant for drives that cannot be connected as SATA. It is not meant to be an adapter for SATA to USB. It could be used for other types of storage devices such as SD cards or NVMe drives with a USB adapter.

I only have SATA drives to test with, so all of my tests are using some form of USB to SATA adapter, which I say you should not be doing. But that is all I have to test with at this time. I will likely be asking for some beta testing help.

My testing with different USB adapters gave different results with different adapters. That means that the real world results will depend on the USB bridge of the device.

For SATA not all adapters directly support passthrough soft or hard resets, so I would think this could carry over to actual USB only drives.

Even without direct support for resets, there is a special reset for USB mass storage devices, called a Bulk Only Reset. This reset will trigger whatever resets the USB bridge needs to do (and is capable of) to reset the storage device. For SATA this reset itself will likely trigger a soft or hard reset, if it is capable. In the event of other types of adapters, it should be possible it would attempt to perform whatever reset could be done.

Resets can be performed on a timeout setting. If the device has not responded within the timeout limit, the reset process is started. If it is successful, the device will become ready again for the next command. If the reset is not supported or has no effect, you then still have to wait for the device to become ready.

In my testing, I have successfully used short timeouts to reset USB connected SATA drives to speed up the error handling process, very similar to timeouts on a SATA connected drive.

For drives that have some sort of “slow” issue, it can be possible to speed up the overall read speed by greatly increasing the cluster size (which is not possible using the OS reads). The idea is that the drive is performing a background operation between reads which causes all reads to take a long time. By increasing the read size, the ratio of data read per background operation is increased, allowing for a much faster overall data transfer. I only have one WD drive to test this with, but I am able to increase the overall read speed by a factor of 100.

For devices that could potentially require a very long timeout setting, you can set the timeout up to a stupid high 15 minutes (same as the other direct modes).

For devices that need to be power cycled, it will work with an external relay (which you would need to purchase and wire on your own, the instructions are on the website). Along with the usual drive issues, this may work well for some solid state devices that need to be power cycled often on a short timeout.

Overall, the Direct USB mode has good potential, but it still relies on the USB bridge, so results will vary. Only further testing will give a sign of the true potential.

 

[Jared]

Keep up the great work!

 

[drdoc]

cloning 520 byte sector drive.

I have a 520 byte sector drive. I am making an image file.
I nneed to strip the first or last 8 bytes from each sector to
make it 'normal'.
Is there a way to do that ?
Is there a script someone already wrote?
I don't see it built in to superclone.

I can write a c program to do it - or maybe through dd....

 

[maximus]

cloning 520 byte sector drive.

I have a 520 byte sector drive. I am making an image file.
I nneed to strip the first or last 8 bytes from each sector to
make it 'normal'.
Is there a way to do that ?
Is there a script someone already wrote?
I don't see it built in to superclone.

I can write a c program to do it - or maybe through dd....

Strait from the HDDSuperClone user manual:

To clone a 520 byte drive to a 512 byte drive (or image file) requires changing the advanced setting “Output sector size adjustment”. To clone from a 520 byte disk to a 512 byte disk, you would change this setting to negative 8. This will trim the last 8 bytes from the 520 byte sectors to achieve 512 bytes.