Two instances of an application can start DJM with same nodeId .
Application A:
val djm = DistributedJobManager(..., nodeId=7,..)
Application B:
val djm = DistributedJobManager(..., nodeId=7,..)
This could prevent jobs from start as as a best case scenario and dramatically obfuscate situation with DJM Zookeeper state in the cluster.
We should prevent DJM from starting if there are another alive DJM instance with same node id.
This will protect our DJM Zookeeper state from being messing around by two DJM instances.