Symptom
FTD VMWare: High disk utilization on /dev/sda8 partition caused by file system corruption
Conditions
-VMWare
-ungraceful shutdowns or bad SSD
Workaround
1. Keep a snapshot of the VM as a precautionary step before proceeding.
2. Login to the system (console/ssh), then go to expert mode and run 'sudo -i' to get to the root user.
3. Run 'pmtool shutdown' to shutdown all FTD processes.
4. Mark the filesystem as dirty - run : 'sync' followed by 'debugfs -w /dev/sda8 -R dirty'
5. Check the state of the file system: 'tune2fs -l /dev/sda8 | grep -i "Filesystem State"'
It should say: "Filesystem state: not clean"
6. From VMWARE side, do a power reset of the VM. That should force a FSCK during the next bootup.
6. FTD VM boots up - Watch for error while it is doing fsck on partition /dev/sda8.
8. Login and run 'df /dev/sda8' to see if fsck successfully recovered the blocks.
Further Problem Description
-The core issue appears to be CSCwc41661 present in 7.0.4. That causes process_stdout/err.logs etc. keep growing over time written over a deleted file handle. A graceful reboot typically recovers from this situation. However, if the box is ungracefully shutdown, it’s possible that the fsck was unable to recover all the free blocks next time it runs during boot up.
