[docs] Add Production Readiness Checklist

[hanliu0830]

Search before asking

• I searched in the issues and found nothing similar.

Description

Follow-up to #4335.

Add a Production Readiness Checklist page under Maintenance > Operations, starting with ZooKeeper deployment and configuration recommendations for Fluss.

The page should cover:

• ZooKeeper's role in metadata storage, cluster coordination and configuration management, a recommendation to use ZooKeeper 3.8 or later, and links to the official download and deployment documentation.
• Single-node development/testing and 3/5/7-node production ensembles with their failure tolerance, placement across failure domains, a 1 vCPU/4 GB starting point, and persistent storage with transaction-log I/O isolation.
• Separate conf/zoo.cfg and conf/java.env examples with parameter descriptions, including ensemble membership, timeouts, data directories, four-letter commands, and automatic cleanup of ZooKeeper snapshot and transaction log files.
• Why Fluss metadata can exceed ZooKeeper's default 1 MiB limit, and how to configure jute.maxbuffer=104857600 (100 MiB) on every ZooKeeper server to match the default Fluss zookeeper.client.max-buffer-size.

Willingness to contribute

• I'm willing to submit a PR!