I have a problem with my SSD drive – it would appear that the controller started acting up. What happens is that if requested to read specific sectors, it will completely shut down the access to disk. At this point not even restarting or ACPI power off helps – I actually need to physically unplug and plug the cord back to make it work again.
I naturally want to rescue as much as possible, but it’s extremely time consuming with ddrescue, given the way it works. I played with –i option, trying to force it to skip specific areas of the disk, but either I do not fully understand the way it works, or the tool ignores the option most of the times. Mind that I do understand that “–i” by design may be ignored depending on the direction ddrescue attempts to read sectors from, yet many times it was ignored despite the right order of reading - e.g. in case my “–i” byte was X+1000 but ddrescue upon starting still attempted to read byte X in “forward” direction.
My question here is whether there is a way to mark some byte “untouchable”, so that ddrescue avoids reading it altogether - possibly by manually editing the log file? If not, do you have any other tips on how to recover the data efficiently?
EDIT: just found Mapfile (logfile) documentation and based on what I see there, it should be fairly easy to prepare it to make ddrescue do what I want it to.