http://bugzilla.novell.com/show_bug.cgi?id=571841 http://bugzilla.novell.com/show_bug.cgi?id=571841#c3 Martin Mohring <martin.mohring@5etech.eu> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |martin.mohring@5etech.eu --- Comment #3 from Martin Mohring <martin.mohring@5etech.eu> 2010-06-20 23:20:14 UTC --- I had even more extreme situation. I was using dump (with read only access to the block device) and had created an ext4 fs on a LVM volume. Then I have just tried to check what happens when I do backup using dump. To my surprise: 0. After a while, I get a read error on one of of the Volumes 1. The LVM Volume Group gets destroyed 2. All ext4 filesystems are then lost The only way to get rid of such a LVM is to null it with dd and putting lots of 0 bytes into the underlying partitions. My only explaination for this is that dump somehow tries to read bytes at non existing fs offset. This causes a non recoverable read error, destroying then the mounted filesystem. I though this could never ever happen, because logically dump accesses only with read access. This can only be explained with LVM bug and/or missing error handling in ext4. So keep off of using dump on a ext4 inside a LVM volume. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.