Distributed Patterns¶

Tier: 1 (System Constraints)

Concepts: 13

Durability: Timeless (Patterns emerge from physical constraints of network communication: latency, partial failure, message ordering)

Fundamental Tensions¶

Tension	Description
Consistency vs. Availability	The CAP theorem shows that under network partition, a system must choose between serving stale data (availability) or refusing service (consistency). This is a mathematical constraint proven by Brewer and formalized by Gilbert/Lynch.
Autonomy vs. Coordination	Independent services can fail/recover independently but require complex protocols for cross-service operations. Tight coordination (2PC) blocks on slow participants; loose coordination (sagas) requires compensation logic.
Latency vs. Correctness	Faster responses often mean less certainty. Waiting longer for acknowledgments increases reliability but degrades user experience. This is bounded by physics (speed of light) and probability (failure detection).
Simplicity vs. Resilience	Each resilience pattern (circuit breakers, retries, fallbacks) adds complexity. Simple systems are easier to reason about but fail catastrophically; resilient systems survive failures but are harder to debug.
Throughput vs. Ordering	Parallel processing increases throughput but loses ordering guarantees. Sequential processing preserves order but creates bottlenecks. This tension is fundamental to parallelism itself.

This Domain Concept	Transfers To
Idempotency	API Design, Event Processing, Payment Systems, Database Design
Eventual Consistency	Database Design, Caching, Distributed Filesystems, Collaboration Software
Saga Pattern	Workflow Systems, Payment Processing, Order Fulfillment, Travel Booking
Circuit Breaker	Electrical Engineering, API Clients, Service Meshes, Mobile Apps
Backpressure	Network Protocols, Stream Processing, Database Connections, Message Queues
Exponential Backoff	Network Protocols, API Clients, Database Connections, Distributed Locks
Timeout Design	Operating Systems, Database Queries, User Interfaces, Real-time Systems
Bulkhead Isolation	Ship Design, Operating Systems, Cloud Architecture, Container Orchestration
Consistent Hashing	Caching, Load Balancing, Database Sharding, Content Delivery
Sharding/Partitioning	Database Design, Multi-tenancy, Search Systems, Event Streaming
Leader-Follower	Database Replication, Distributed Consensus, Message Queues, Distributed Locks
Outbox Pattern	Event Sourcing, Message Queues, Data Integration, Audit Logging
Load Shedding	Operating Systems, Electrical Grids, Triage Medicine, Airport Security

Source	Coverage
Kleppmann, "Designing Data-Intensive Applications" (2017)	Comprehensive patterns synthesis; replication, partitioning, consistency models
Nygard, "Release It!" (2007, 2018)	Resilience patterns: circuit breaker, bulkhead, timeout
Helland, "Life Beyond Distributed Transactions" (2007)	Saga rationale; practical distributed system design without 2PC
Hohpe & Woolf, "Enterprise Integration Patterns" (2003)	Messaging patterns; integration architecture
Garcia-Molina & Salem, "Sagas" (1987)	Original saga pattern paper
Karger et al., "Consistent Hashing and Random Trees" (1997)	Foundational consistent hashing paper
DeCandia et al., "Dynamo: Amazon's Highly Available Key-value Store" (2007)	Eventually consistent systems; consistent hashing in practice
Vogels, "Eventually Consistent" (2008)	Consistency model clarification
Brewer, "CAP Twelve Years Later" (2012)	CAP theorem nuances
Lamport, "Paxos Made Simple" (2001)	Consensus fundamentals
Ongaro & Ousterhout, "In Search of an Understandable Consensus Algorithm (Raft)" (2014)	Accessible consensus algorithm
Google SRE Book (2016)	Operational patterns: load shedding, timeout budgets
Richardson, "Microservices Patterns" (2018)	Modern saga and outbox implementations
Jacobson, "Congestion Avoidance and Control" (1988)	Foundational network flow control
Welsh et al., "SEDA" (2001)	Staged event-driven architecture; load shedding