HiveEndPoint, HiveStreamingConnection: Optimise metastore calls to open transactions batch and allocate write Ids.

Per table write ID implementation (HIVE-18192) have introduced write ID and maps it against the txn.

Now, for streaming ingest, we need to open txns batch and then allocate write id for each txn in the batch which is 2 metastore calls.

This can be optimised to use only one metastore api.