Hi All,

I have few doubts on File System. Those doubts are:

In modern file systems like GFS, one DINODE (disk inode) take up a complete data block. Idea behind that is to not put a limit on number of files i.e. one can create files till all the data blocks are not used.

But my doubts are, GFS still uses dinode bitmap as well as data block bitmap.

Q1>. This dinode bitmap, maps to which portion of the disk? Are they map to all the data blocks of the Resource Group of the disk partition?

Q2>. The data block, that contains the dinode, also holds the dinode number. Whereas in UNIX file system dinodes are indexed into the inode list. By checking the dinode number in the data block while accessing the file, does it not hamper the performance?

If any body has any idea on advanced file system like GFS, please help me to understand the file system architecture. Any link will be also helpful for me.


Thanking U all.

With Regards
RajSun.