Adding function toRedisByteZSET

[bowen-dd]

Hi team,

The current implementation of toRedisZSET(kvs: RDD[(String, String)]) is useful for simple use cases but is not sufficient for high-throughput or binary-safe workloads.

We’d like to propose support for a new method:
toRedisByteZSET(kvs: RDD[(Array[Byte], (Double, Array[Byte]))])
This enables writing Redis ZSETs using byte array keys and members, which is critical for:

• Binary-safe use cases (e.g., Protobuf, embeddings, hashed IDs)
• Avoiding costly string serialization and base64 encoding

We’ve implemented this function following the same pattern as toRedisByteHASHes, including host-aware batching and pipelined writes. We'd appreciate it if the team could review and prioritize this addition.

Thank you!