file system check software

[Edward Ned Harvey]

> So basically the test that I need to do is fill up the 6 terabytes of
> the file system with files, large and small and do md5sum checks of the
> files as they are created and copied about on the file system. And do
> this file creation, replaction and deletion over and over again,
> stressing not just the software, (kernel, drivers, etc.) but also the
> underlying hardware, (disks, controllers, cables etc.) Such a tool
> would
> pretty much validate a disk system so that a bank would trust it to
> hold
> its database of customer accounts.

If a file was open for writing at the time while it was being copied, then
the original and the copy might differ from each other.

Most likely you don't need to do anything so complicated.  Since you've
experienced a problem, and you want to identify the cause of the problem, I
bet you can start with the simplest thing:  Just run surface scan on each of
the disks.  Most likely you'll find a failing disk.

If you want to verify that the filesystem is able to read and write
everything it says it can, you might try something really simple, like this:
	dd if=/dev/zero bs=1024k | tee /some/file | md5sum -
	md5sum /some/file
	The two outputs should match.
To make it a little more challenging:
	dd if=/dev/urandom bs=1024k | tee /some/file | md5sum -
	md5sum /some/file
	The two outputs should match.

If you can't do the above, you know you have some failing component, and
you've got to figure out which one (or ones.)  If the above completes
successfully ... That's pretty strong confidence in the hardware ... but for
your purposes maybe you care about testing it more extensively.