5 replies
by cpfeiffe
Your system should be as quite as possible (single user mode). Then you should use ufsdump against the raw device that represents each file system (do one file system at a time). I don't know what OS you are using, but here is a Solaris example:Run 'df -k' to find the file system devices (They will start with /dev on the left. Record these devices (you only need the first column of the output).
Run 'init S' to go to single user mode.
One device at a time, run 'ufsdump 0f FILE DEVICE' where DEVICE is the device you recorded (but replace the dsk portion with rdsk for raw device). and FILE is the output location of the dump. Dumps can be very large so make sure there is enough space where you are dumping to. Some dumps can be as large as the file system (used space) itself.
Good luck.
by bamu
I imagine that you want to fix the error on this disk. First you need to umount that filesystem, just to be sure that no one will access it will you are performing the backup. then issue that command:
ufsdump 0f tapename /dev/rdsk/filesystemif there is some error on that disk and you want to fix it:
fsck -Y /dev/rdsk/filesystem
The -Y will answer YES to all questions, if you want to see all questions, just remove the -Y.
If there is still some errors after the fsck, you might try one of these 2 :
- newfs on that device and restore your backup
on it.
or
- if you have a spare disk, you could format that disk like the original, newfs on each partition, mount that disk, and copy the defect disk on the new one with that command:
ufsdump 0f - /dev/rdsk/filesystem|(cd /newdisk;ufsrestore -xf - )
by bamu
by bamu
This discussion is closed to new replies.
Trending Linux posts
Sort by Trending