[solved]Recovery of Linux partions
Posted: 18 Jul 2014, 09:13
Hello,
at the moment I try to recover a Linux system after a hdd failure. After a system update the machine did not reboot so I turned it off manually. After that it could not boot anymore, so I got a systemrescuecd and had a look. In gparted the hdd showed up as completly unlocated. So I gave testdisk a try, the scans showed an error at an low cylinder count (~5800 of ~120k). After accidently writing a new mbr, I thought it would be better to make a copy. So I used ddrescue to clone the drive on to an identical one. ddrescue could copy the hole disk, except the damaged part which is only ~1.5MB (the disk has 1 TB).
I now used tesdisk to scan the new disk. Since the result with "intel PC partion" made no sense and I could not see any folders. I also tried "None" and "Efi GPT". For all options I did also a "deep scan". Now I am totally confused what to do. The result from the "None" not-deep-scan, where the most reasonable (3 Partitions: a small ext3/4, a big ext3/4 and swap). There I could also access the files, but almost everything is missing.
Due all the scans the logfile got a bit big, so I put it here for download: https://www.dropbox.com/s/hros758j9t4t9m0/testdisk.txt
If you don't want to download please point me to some parts that could help and I try to copy-paste them.
Just for clarification sda is the faulty drive, sdb is the drive with the copy of sda. Both are Seagate 1TB with 4k sectorsize but are emulating 512k. The structure I am looking for is a ~100GB ext partition for boot and root, a ~890 GB for /home and a ~8GB for swap. If I remember right there should be also a very small unlocated part.
Sorry for the lengthy text. I hope you can help with that.
Thanks in advance!
at the moment I try to recover a Linux system after a hdd failure. After a system update the machine did not reboot so I turned it off manually. After that it could not boot anymore, so I got a systemrescuecd and had a look. In gparted the hdd showed up as completly unlocated. So I gave testdisk a try, the scans showed an error at an low cylinder count (~5800 of ~120k). After accidently writing a new mbr, I thought it would be better to make a copy. So I used ddrescue to clone the drive on to an identical one. ddrescue could copy the hole disk, except the damaged part which is only ~1.5MB (the disk has 1 TB).
I now used tesdisk to scan the new disk. Since the result with "intel PC partion" made no sense and I could not see any folders. I also tried "None" and "Efi GPT". For all options I did also a "deep scan". Now I am totally confused what to do. The result from the "None" not-deep-scan, where the most reasonable (3 Partitions: a small ext3/4, a big ext3/4 and swap). There I could also access the files, but almost everything is missing.
Due all the scans the logfile got a bit big, so I put it here for download: https://www.dropbox.com/s/hros758j9t4t9m0/testdisk.txt
If you don't want to download please point me to some parts that could help and I try to copy-paste them.
Just for clarification sda is the faulty drive, sdb is the drive with the copy of sda. Both are Seagate 1TB with 4k sectorsize but are emulating 512k. The structure I am looking for is a ~100GB ext partition for boot and root, a ~890 GB for /home and a ~8GB for swap. If I remember right there should be also a very small unlocated part.
Sorry for the lengthy text. I hope you can help with that.
Thanks in advance!