I (partly) manage a small cluster and we are having some problems on a few nodes (running CentOS 6.4). Some processes show exceptionally long run-times in 'top', even though the computer has been rebooted very recently. See here for an example:
Note - up time only 2 days, 'init' process time almost 4 million hours.
The problems maybe started after the clocks changed but the system time seems to be correct, as shown at the bottom of the screenshot.
Any suggestions on what might be causing this, and more importantly how to fix it?
Edit - ps: