A multi-region configuration distributes Kafka components across separate geographic cloud or data-center regions. It addresses resilience when a complete data-center outage would otherwise interrupt service, while introducing replication latency, networking costs, and data-consistency challenges. section 9.2.1
Evaluate the pattern against the operational conditions it creates:
- Metadata availability depends on a majority of Kafka servers, including KRaft controllers or ZooKeeper nodes, remaining operational. section 9.2.1
- Surviving a complete data-center outage requires clusters to span at least three data centers; smaller organizations may accept outage risk when that infrastructure is unavailable. section 9.2.1
- Cloud infrastructure makes cross-region broker distribution more feasible and reduces operational burden, but does not remove replication, cost, or consistency trade-offs. section 9.2.1
- Deployment selection still balances operational complexity, scalability, fault tolerance, security, and compliance across on-premises, cloud, and hybrid environments. section 9.2
Cloud-based regional distribution
Distribute Kafka brokers across multiple cloud regions to improve resilience, then assess whether replication latency, network charges, and data consistency remain acceptable for the deployment's requirements. section 9.2.1
