- Oct 5, 2016
- 1
- 0
- 1,510
Short version: With a HDD with a bad sector and IO errors, I want to use an Ubuntu live usb to ddrescue what I can, but Ubuntu is taking forever to start up. Is there a way I can push it to ignore the hard drive problems on mounting so that I can just make a raw drive image and be done with it?
Long version: I have an 80GB IDE HDD that was once installed in an older computer.
I am not sure what is on it, but I'd like to find out before I trash it if I can. I know sometimes what is lost is lost, and I'm usually pretty good about backing up before I retire a drive or system, so I am not in a panic. I just want to know what I can do and when to let go.
So. I have an Acomdata usb HDD enclosure and have put the drive in it, and connected to my new pc.
Windows 10 makes the new usb sound, and after about an eternity a drive letter H pops up. So it appears this has been assigned a letter before, though it could just be from the last time I tried to look at it and quit on it when I realized there was a problem.
Clicking on H eventually causes Windows to prompt me to format. Of course I didn't.
I remember disk management showing the drive (or partition?) as raw at some point and that was when I put the brakes on.
Now I have some deep scan recovery software and am back at it for one more go.
I tried connecting it to my Mac and it comes up as uninitialized. When I ignore, I can look at the disk info and see that there is a lost NTFS partition.
I went back to Windows because that is where I have the recovery prog installed. I began to scan the disk. It was taking longer than usual to assess the disk, so I walked away.
But after coming back, the prog said the disk had disconnected. I figured there was a usb power setting that set it to sleep, so I set everything on my computer to performance power settings and tried again.
I came back and once it did its preliminary assessment, it estimated 300+ hours to scan. At first I thought I'd leave it and just let it do its thing, and once it got past whatever hiccup there is it would finish quicker.
But after some reading, I stopped it right there and realized that there must be a physical problem, and decided I need to image the disk and recover from the image.
I tried to look at the SMART data with crystaldiskinfo but the disk doesn't show up. So I don't know what that says yet, but I am proceeding as though there is a physical problem. There are no odd sounds coming from it though. My only other indicator that there is a problem with it is that the power light on the enclosure is usually blue, but it goes red with this disk.
I tried to use dd on my Mac, but it got a short way into it and threw an IO error.
THEN, it occurred to me that I should see what the jumper is set to. I haven't dealt with pata drives in quite a while now. There was no jumper, which was the slave setting on this drive. I put a jumper on, but went through the "which way is up?" thing when looking at the diagram. I put the jumper on what (I guessed) was the master setting and tried again.
It powered up fast and I had a blue light. Cool. The system recognized it as a 31.whatever GB drive and I knew that I had it backwards and put the jumper on the "limit to 32GB" setting.
Well at least I know the damaged sector is on the latter half of the drive...
I put the jumper back to master just to see if I can image the whole disk in Ubuntu.
So now I am booting from an Ubuntu live USB and plan to run testdisk for some more info and hopefully ddrescue. But on booting I got a few blk_update_request IO errors, which all indicate the same sector. Also a buffer IO error, async page.
<strike>And now I am staring at ubuntu's purple screen and blinking dots.</strike>
[Edit to add: I clicked the mouse and the purple screen went away and revealed the activity. It is repeating the errors with various sector and block numbers. Not sure if I should wait or jump ship...]
It is clearly stuck on trying to handle this drive. Is there a way to push ubuntu to just let me access the drive without it working on sorting out the errors? I just want to image the good blocks and skip the bad blocks, take what I can and then take all my old hard drives to the shooting range.
I know that failing that, I can set the drive to 32GB and recover off that part of the drive.
If you read this far, I am grateful. What do you think about imaging the whole disk, including the damaged half?
Thanks
Long version: I have an 80GB IDE HDD that was once installed in an older computer.
I am not sure what is on it, but I'd like to find out before I trash it if I can. I know sometimes what is lost is lost, and I'm usually pretty good about backing up before I retire a drive or system, so I am not in a panic. I just want to know what I can do and when to let go.
So. I have an Acomdata usb HDD enclosure and have put the drive in it, and connected to my new pc.
Windows 10 makes the new usb sound, and after about an eternity a drive letter H pops up. So it appears this has been assigned a letter before, though it could just be from the last time I tried to look at it and quit on it when I realized there was a problem.
Clicking on H eventually causes Windows to prompt me to format. Of course I didn't.
I remember disk management showing the drive (or partition?) as raw at some point and that was when I put the brakes on.
Now I have some deep scan recovery software and am back at it for one more go.
I tried connecting it to my Mac and it comes up as uninitialized. When I ignore, I can look at the disk info and see that there is a lost NTFS partition.
I went back to Windows because that is where I have the recovery prog installed. I began to scan the disk. It was taking longer than usual to assess the disk, so I walked away.
But after coming back, the prog said the disk had disconnected. I figured there was a usb power setting that set it to sleep, so I set everything on my computer to performance power settings and tried again.
I came back and once it did its preliminary assessment, it estimated 300+ hours to scan. At first I thought I'd leave it and just let it do its thing, and once it got past whatever hiccup there is it would finish quicker.
But after some reading, I stopped it right there and realized that there must be a physical problem, and decided I need to image the disk and recover from the image.
I tried to look at the SMART data with crystaldiskinfo but the disk doesn't show up. So I don't know what that says yet, but I am proceeding as though there is a physical problem. There are no odd sounds coming from it though. My only other indicator that there is a problem with it is that the power light on the enclosure is usually blue, but it goes red with this disk.
I tried to use dd on my Mac, but it got a short way into it and threw an IO error.
THEN, it occurred to me that I should see what the jumper is set to. I haven't dealt with pata drives in quite a while now. There was no jumper, which was the slave setting on this drive. I put a jumper on, but went through the "which way is up?" thing when looking at the diagram. I put the jumper on what (I guessed) was the master setting and tried again.
It powered up fast and I had a blue light. Cool. The system recognized it as a 31.whatever GB drive and I knew that I had it backwards and put the jumper on the "limit to 32GB" setting.
Well at least I know the damaged sector is on the latter half of the drive...
I put the jumper back to master just to see if I can image the whole disk in Ubuntu.
So now I am booting from an Ubuntu live USB and plan to run testdisk for some more info and hopefully ddrescue. But on booting I got a few blk_update_request IO errors, which all indicate the same sector. Also a buffer IO error, async page.
<strike>And now I am staring at ubuntu's purple screen and blinking dots.</strike>
[Edit to add: I clicked the mouse and the purple screen went away and revealed the activity. It is repeating the errors with various sector and block numbers. Not sure if I should wait or jump ship...]
It is clearly stuck on trying to handle this drive. Is there a way to push ubuntu to just let me access the drive without it working on sorting out the errors? I just want to image the good blocks and skip the bad blocks, take what I can and then take all my old hard drives to the shooting range.
I know that failing that, I can set the drive to 32GB and recover off that part of the drive.
If you read this far, I am grateful. What do you think about imaging the whole disk, including the damaged half?
Thanks
Facebook
Twitter
Instagram