Backend Engineering

Backend engineering is the central track of this handbook. It covers the design and operation of APIs and services that remain correct under concurrency, retries, partial failures, and changing load.

HTTP and API design

• HTTP semantics, idempotency, and pagination
• API authentication and authorization — identity, OAuth/OIDC, token validation, object-level access control, request validation, and API compatibility
• Error contracts and backwards compatibility

Data and persistence

• SQL indexes, query plans, transactions, and isolation
• Database connection pooling for backend services — deployment-wide connection budgets, bounded pools, transaction pooling, and saturation diagnosis
• Application caching: consistency, invalidation, and stampede protection
• Zero-downtime database schema migrations — expand/contract, compatible releases, batched backfills, and staged constraints

Distributed backend systems

• Messaging, delivery guarantees, retries, and dead-letter queues
• Microservice communication and resilience
• Distributed rate limiting for multi-tenant APIs — algorithms, atomic shared state, tenant fairness, HTTP rejection semantics, and failure modes
• Distributed leases and fencing tokens — owner identity, monotonic generations, stale-writer rejection, database coordination, and leader election
• Backpressure, admission control, and load shedding
• Observability, SLOs, and incident response

Each topic should connect API and implementation choices to correctness, security, performance, operational failure modes, and maintainability.