Recently I was using my external NTFS-formatted 3TB HDD on my Linux OS again, as I always do. Nothing out of the ordinary.
At some random point I received an Input/Output error when trying to access files. I tried to do other things on the drive, checked more files and it was really slow and it felt like I was dealing with the last living hours of my almost dead hard drive. Therefore I started to rsync my newest changes to the backup drive. Sadly I had to restart the computer at some point which led me to rsyncing once again, but this time it got stuck on a certain folder on the drive which gave Input/Output error again as the output of rsync's error log. I reran rsync more and more and it always got stuck on this folder.
Before I tried to rsync, after I realized something is wrong with the drive I wanted to smartctl it, but it always failed with
Read Device Identity failed: scsi error medium or hardware error (serious)
A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options.
which gave me even more the impression that something was seriously wrong with my HDD. Again, smartctl (with different arguments each time) didn't work for me, at all and it seemed like my HDD is still barely breathing and I wondering about what to do next before it completely dies off. (Note: SMART worked in the past 100%. It is also enabled in BIOS.)
Now that happened all on LINUX. On a Debian testing based derivative, to be more precise.
So let us change to Window's view now:
At some point in the middle of the aforementioned sequence of events I changed to Windows again and ran a CHKDSK, which repaired a couple of errors. Re-running this didn't find any new errors. Also a couple of months ago before this happened I knew that this HDD received exactly 600 uncorrectable sectors. I monitored them and it always stayed at exactly this number.
Now what did Windows say after all that happened?
I rebooted into Windows to check if at least Crystal Disk Info works or if one of the NTFS-native Windows tools could help me out better.
Oh wonder, Crystal Disk Info actually worked. Without even taking long or anything like that. And even more astounding: the only thing wrong with my hard drive was still the exact same amount of 600 uncorrectable sectors. I was expecting some dead piece of metal being in the deep reds of SMART-ville while gaining hundreds of new uncorrectable sectors each minute.
But that didn't happen.
In fact on Windows (almost) everything went fine. The HDD wasn't slow. I could do file operations with it, as usual, etc. The only problem that actually migrated (partially) is the folder from earlier that gave an Input/Output error when trying to rsync. I couldn't even manually delete or move it when in Linux. But hero Windows 10 saved the day by letting me actually delete this whole folder after archiving it. It took a while, but it worked.
So here are my questions:
- What could the possible reasons be for this to happen?
- Why is Linux telling me that the hard drive is so broken that I can't even SMART it while a Windows program can?
- Why is the hard drive very slow and annoying when booted into Linux but working pretty fine on Windows?
I want to understand this...
No comments:
Post a Comment