(TestCase) Agent crashes when Master Scheduler process is killed

Pre Test

1. Setup a JobScheduler / Agent configuration for v1.7.3
2. Start both JobScheduler (Master & Agent)
3. Get the test configuration from JS-1131 and put it in the live folder of the master
4. Change to process-class into the sample pointing to the running agent
5. start the test job1_remote
6. stop master JobScheduler process will "kill" (e.g. via the task manager) => master JobScheduler will be killed
7. be sure that the agent is killed, too.

Pre Test

1. Setup a JobScheduler / Agent configuration for v1.7.4
2. Start both JobScheduler (Master & Agent)
3. Get the test configuration from JS-1131 and put it in the live folder of the master
4. Change to process-class into the sample pointing to the running agent
5. start the test job1_remote
6. stop master JobScheduler process will "kill" (e.g. via the task manager) => master JobScheduler will be killed
7. the agent should always be running.

Expected behaviour

1. the agent should never be killed in case of killing the master
2. after restart of the master the job can be started again in a normal way