Implement Export/Import dataflow

[tuan08]

1. Create a shell command:

shell.sh dataflow export  --dataflow-id  --config-file

This command will use registry.get("/neverwinterdp/scribengin/dataflow/dataflow-id") data.

1. Convert the data to DataflowDescriptor object , using JSONSerializer.
2. get the latest data config in configuration and update the DataflowDescriptor. This is not ready yet, and I need to update some registry code
3. Write the DataflowDescriptor to --config-file file

[KingOfPoptart]

Use case 1:
Entire cluster shuts down cleanly (including kafka, zookeeper, Scribengin, hadoop, etc)
Entire cluster starts back up
Everything should continue as normal

Use case 2:
Export configuration of currently running cluster
Stop the cluster
Start the cluster again with the exported configuration.

Use case 3:
Export configuration of currently running cluster
Stop the cluster
Start a completely new cluster with the exported configuration.