More uniform BGP router ID selection for IPv4 and IPv6

[antoninbas]

Currently, the Node IP will be used as the BGP router ID unconditionally when the Node has an IPv4 address (IPv4-only or dual-stack clusters). For IPv6-only clusters, we will honor the node.antrea.io/bgp-router-id Node annotation (if provided by the user), or fallback to a "randomly-generated" value (in reality, hash-based).

This implementation doesn't make much sense to me. I see no reason why the annotation should not be honored in the IPv4 case. If anything, that just creates confusion. RFC 6286 relaxes the definition of the BGP router ID, and I don't think we need to "force" the ID to be the Node IP in the IPv4 case.

Originally posted by @antoninbas in #6524 (comment)

[rajnkamr]

Originally, the BGP Identifier needed to be a unique identifier for each BGP router. Typically, it was derived from the IP address of the router i.e Node IP and the identifier had to be unique across the entire network,

However now it is defined as a 4-octet (32-bit), unsigned, non-zero integer.
Now Instead of requiring global uniqueness, the identifier only needs to be unique within the same Autonomous System (AS).