HDDSuperClone vs ddrescue
- Thread starter jol
- Start date
regarding ruling out writing process, will let you know when done
regarding the computer specs and linux OS, I tried various computers some better some not also different OSs and versions including your Lubuntu ISO from your website, and there was not any significant differences
the leading line was the same everywhere, which is a diff. between ddrescue and HDDsuperclone about 30% (in favor of ddrescue)
regarding the computer specs and linux OS, I tried various computers some better some not also different OSs and versions including your Lubuntu ISO from your website, and there was not any significant differences
the leading line was the same everywhere, which is a diff. between ddrescue and HDDsuperclone about 30% (in favor of ddrescue)
If it does turn out to be the writing process, please add the option --odirect to the ddrescue command to see if it slows down to the speed of hddsuperclone.
If it is not the writing process, then try running hddsuperclone in the two different modes SCSI and ATA passthrough, and see if there is a difference between them. If that is not it, then add --idirect to the ddrescue command to see if it slows it down to the speed of hddsuperclone.
If it is not the writing process, then try running hddsuperclone in the two different modes SCSI and ATA passthrough, and see if there is a difference between them. If that is not it, then add --idirect to the ddrescue command to see if it slows it down to the speed of hddsuperclone.
After some testing I have found the condition that causes this. Most of my testing has been to image files which seem to respond better to using a direct non-buffered write. But when cloning to another disk the non-buffered writing is slower, by almost exactly the 30% reported. So my instinct that it is the writing process is correct. I will add an option in the next release to address this. I still have to figure out how best to implement it since it seems better one way or the other if the destination is an image file or a drive.
I have added an option in the GUI to enable or disable the OS write buffer in the latest release (version 1.10 released 8-27-17). It is enabled by default, which seems to be the faster setting. For those still using the command line mode, it is enabled by default with no way to disable it.
maximus":2b1nrib2 said:
Please test with this latest version to confirm that it resolves the major speed difference.
I would like to point out that in my testing, I did find that for some reason, the ATA passthrough mode seems to be slower than the SCSI passthrough mode in some cases. It seems to be just as fast until it hits bad sectors, which does not make any sense considering it is a more direct way of reading.
Well, it is at least changed in the command line mode to be enabled, so it should be faster now. And be prepared, I now have all the free options in the GUI. Once I get all the other special options in the GUI, the command line mode will be removed. Maybe not right away, but I am not adding any new command line options. It is too difficult to maintain both the GUI and command line.