How does the 'Cooperative Sticky' rebalancing protocol improve upon the legacy 'Eager' protocol in large clusters?

1. A It allows the broker to skip the ZooKeeper metadata check entirely for every third rebalance.
2. B Incremental Revocation: Instead of a 'Stop-the-World' pause where all members drop all partitions, it allows members to keep their current assignments and only revokes the minimal set of partitions that actually need to move, maintaining higher throughput.
3. C It forces the Producer to pause all writes until the Consumer group is perfectly stable again.
4. D It encrypts the group coordinator's state using a shared secret between all members.