Top 7 Most-Used Distributed System Patterns

Ambassador: Intermediary between app and services. Offloads logging, monitoring, retries. Reduces latency, enhances security. Tech: Kubernetes, Envoy.
Circuit Breaker: Prevents cascading failures. Stops requests to unavailable service, allows recovery. Tech: Netflix Hystrix.
CQRS (Command Query Responsibility Segregation): Separates write operations (commands) from read operations (queries). Scales, optimizes independently. Good for different performance, resource needs.
Event Sourcing: Stores ledger of events representing changes instead of direct record updates. Full history, auditing, debugging, Time Travel Debugging (event replay). Tech: Git.
Leader Election: Ensures single node active for specific task. Elects new leader on node failure. Avoids conflicts. Tech: ZooKeeper, etcd.
Pub/Sub (Publisher/Subscriber): Asynchronous messaging. Publishers emit events. Subscribers listen. High scale, modularity. Propagates updates across services. Tech: Google Cloud Pub/Sub.
Sharding: Horizontally distributes data subsets across multiple nodes. Improves performance, scalability, data locality, query speed. Tech: MongoDB, Cassandra.
Strangler Fig (Bonus): Gradual legacy system replacement. Incremental component migration. Avoids high-risk big bang migrations.

More in this category