File System : Block Groups
Hi, I have a query regarding the purpose of dividing the entire partition into block groups(as done by ext2 , ext3 ). The main purpose is to replicate the super_block and other important information so that it can be recovered in case of data corruption. Other purpose is to enable efficient storage in order to minimize defragmentation and head seek time. My questions are : Are there any other purpose of this ? How it is possible to detect that there is some corruption in the main super_block information ? Is some kind of check sum is used ? What is the standard way of doing it ? -- With Regards Dibyayan Chakraborty
On Sat, 25 May 2013 23:56:50 +0530, Dibyayan Chakraborty said:
The main purpose is to replicate the super_block and other important information so that it can be recovered in case of data corruption. Other purpose is to enable efficient storage in order to minimize defragmentation and head seek time.
My questions are :
Are there any other purpose of this ?
What, that isn't reason enough? ;)
How it is possible to detect that there is some corruption in the main super_block information ? Is some kind of check sum is used ? What is the standard way of doing it ?
In fs/ext4/super.c, see the various ext4_superblock_csum-* functions and their callers, around line 115 to 160 or so. Figuring out what other filesystems do is left as an exercise for the reader. ;)
participants (2)
-
Dibyayan Chakraborty -
Valdis.Kletnieks＠vt.edu