Do you restart the database?
include in the first line of log_full.sh :
Code:
exec >/tmp/log.log_full 2>&1
set -x
echo $0 $*
Force to next logical log with "onmode -l" and monitor the log of database (onstat -m) and the file /tmp/log.log_full
check all logs and look for any error or misbehavior.