Calling start-shuffle-service.sh script is not necessary and can cause worker registration to fail

[jafreck]

https://spark.apache.org/docs/latest/job-scheduling.html#configuration-and-setup

"In standalone mode, simply start your workers with spark.shuffle.service.enabled set to true."

It seems the external shuffle service is built in to standalone mode. So only the spark.shuffle.service.enabled spark-defaults.conf is required. The script does not need to be run and seems to occasionally cause worker registration to fail as the workers and the external shuffle service both try to listen to port 7337.

[jafreck]

Fixed by #492