Resilience is a system’s ability to continue providing required behavior when components or dependencies fail. In Kafka architectures, it protects business data from component loss through durable storage and replication. section 1.7.3 section 5.2.8
Kafka supports resilience through several architectural choices:
- Replicates partitions across brokers so that, with the correct configuration, at least one copy remains available after a component failure. section 5.2.8
- Persists an immutable event sequence, improving durability while supporting high-throughput distributed processing. section 1.7.3
- Decouples event-driven microservices, improving scalability and reducing the impact of failures between services. section 5.5
Business data that cannot be lost
Financial, healthcare, and e-commerce systems can duplicate components through Kafka partition replication so that one broker failure does not cause data loss. section 5.2.8
Resilience is an architectural trade-off rather than a default guarantee. Deployment decisions must balance fault tolerance with operational complexity, scalability, multi-region design, security, compliance, and infrastructure management. section 9.2
