If you change a job configuration while the database connection get lost then the JobScheduler can crash

The following steps force a JobScheduler crash:

• a job is choosen in JOC and displayed in the right frame
• stop the database while JobScheduler is running
• change a job configuration (e.g. the run_time)
• wait a little time until the JobScheduler knows that the configuration has changed
• click "update" in JOC which sends particularly:

  <check_folders/>
  

• start the database

Then after a while (when the automatic db-reconnect occurs) the JobScheduler is down.
See attached log.