Support load balancing of connection on multiple clusters

[Sfurti-yb]

Problem statement:

In a multi-cluster environment, the current implementation of the pgjdbc driver fails to handle connections correctly. The clusterinfomap does not distinguish between nodes belonging to different clusters. As a result:

• All nodes of all clusters are stored together in the same map.
• When a connection request comes in for a specific cluster, the driver may incorrectly route the connection to a node in a different cluster.

This lack of demarcation for nodes across clusters prevents reliable support for multi-cluster deployments and introduces ambiguity in connection routing.

Changes made:

A new column, universe_uuid, is to be introduced in the existing yb_servers function. This column will serve as a unique identifier for the universe (cluster) to which each server belongs.
Details of the implementation provided in this doc.

Testing

Test cases being added to the driver-examples repo. TODO: Add link here once done