I've got a VMWare machine that I am attempting to test when the command option in /etc/sysconfig/autofsck will really instantly answer yes when fsck operates. I have used debugfs to remove inodes, change block amounts and usually screw using the superblock using the expected lack of some time and errors after i by hand run e2fsck in save mode. However, the errors (aside from the superblock problems which prevent bootup) aren't detected throughout the fsck that's forced at bootup. Using DD to at random write data towards the hard disk does not guarantee that I'll have the ability to recover with fsck (although it does a fantastic job of replicating using the cover off a spinning hard disk and tossing sand inside it).
TLDR: I have to consistently create corruption of the virtual machine's hard disk that may be retrieved using fsck/e2fsck, and I have to get fsck throughout boot as much as recognize it and instantly repair it as well as it may unwatched.
*I am in way over my mind, I understand.