ext4 filesystem corruption -- maybe hardware error?

Solution 1:

First, you might want to do a full fsck of the root disk. Sometimes, I have found that the quick check sometimes misses some important errors. You can do this by either touch a file in the root directory (maybe Linux distribution dependent) but might try

 touch /forcefsck

AND rebooting OR starting up the rescue CD and doing the performing the fsck of the root there. By full, I mean use the -f fsck parameter.

Second, is your syslog indicating any hardware errors?

As Mr. Kario indicated, you might look at checking the disk health using smartctl. I find that some disks that I have used do not report information however.

Solution 2:

198 Offline_Uncorrectable   0x0000   002   107   000    Old_age   Offline      -       21198

It has failed, RMA it.

You may want to run SMART test on it, but with such values it's just a formality, it's highly unlikely it won't fail.

To run a test, use

smartctl -t long /dev/sda

It will tell you when the test will end, then you run smartctl -a /dev/sda again it will show the test result in self test section.