* From the logs, I'm guessing probable cause is that an exception generated by a bad index given to substring error line number conversion stopped the script being killed, leading to continuous events that filled up the log (maybe) * Someone will need to go back and fix this properly