Hello,
At file system(udf disk based) level i need to implement bad block manager. Plz give me some approches, or let me how this implemented in other file systems

Some info, and there is probably lots more

http://packages.ubuntu.com/gutsy/admin/linux-patch-evms

Moved: This thread is more suitable in the kernel forum and has been moved accordingly to help your thread/question get the exposure it deserves. Please note posting duplicate threads is against the LQ Rules so don't do that. Please read the LQ Rules before continuing. Instead use the report button to ask moderators to move your thread to the forum of your choice. Finally *do* respond to replies made, you left your previous thread unanswered which doesn't help anyone.

Sorry. By mistake i had sent in another forum.
Thanks for moving it to proper forum

As the article cited (and many others) note, the general approach is to use hardware-RAID, and then on top of that to use a logical-volume approach which introduces the notion of "storage pools" instead of physical drives. This gives rise to, among other things, the very curiously-named SAN (Storage Area Network), which is a self-managing hardware device specifically dedicated to storage. A less expensive but also less-sophisticated technology is NAS (Network Attached Storage), which more-directly exposes "only its drives."

"Bad blocks," therefore, are dealt-with by avoidance. Run-of-the-mill disk drives support an internal self-test mechanism known as "SMART," which Linux gives direct access to and which storage devices also exploit. A drive that judges itself "pending failure" is retired and taken off-line by the storage controller ahead of time.