belbo wrote:
You can have good practices, to avoid exposing the drive to unnecessary danger, like mechanic impact, eletric fields or bad power supply. All this won't make sure you are safe from badblocks, but will hopefully help. But this is BEFORE the problem. After badblocks appear, theres nothing to do to save the drive, unless you believe in magic, vodoo, freezing the drive and then warming it up and other folk stories... What you can try to do is save whatever data is possible to be saved from it, backing it all up. Do it soon.Bruno Buys wrote:belbo wrote:[...]journal_bmap: journal block not found at offset 5132 on hdb5 Aborting journal on device hdb5. __journal_remove_journal_head: freeing b_committed_data journal commit I/O error ext3_abort called. EXT3-fs error (device hdb5): ext3_journal_start_sb: Detected aborted journal Remounting filesystem read-only If I umount the partition, do e2fsck -y /dev/hdb5 and remount, partition works fine for 2-3 hours, and then it happens again. I think it is only a software problem, what can I do to fix the problem? Thanks and byeLooks like hw failure to me (badblocks?).I did "fsck.ext3 -c /dev/hdb5". It finds many badblock, but it is useless. How can I avoid badblocks? Bye
Good luck!