Temporal Workflows: Durable Orchestration for Microservices and Agents

"Temporal Workflows: Durable Orchestration for Microservices and Agents"

Modern systems don’t fail because you can’t run code—they fail because coordinating code across time, retries, deploys, and partial outages is harder than any single service. This book targets experienced engineers and architects who build microservice platforms, backend workflows, and agentic applications and want a rigorous, practical model for turning ordinary code into durable, crash-resilient orchestration using Temporal.

You’ll develop a precise mental model of Temporal’s execution semantics: event history as source of truth, replay-driven fault tolerance, determinism constraints, and how workers and task queues scale. From there, you’ll learn to design activities as the boundary for side effects, engineer timeouts and retries as explicit contracts, and apply idempotency and compensation to make failures safe. The book covers core orchestration patterns (timers, fan-out/fan-in, child workflows, sagas), robust workflow APIs with signals/queries/updates (including Update-With-Start), and production operations: visibility, search attributes, metrics, tracing, schedules, and CLI-based incident workflows. Finally, it extends these ideas to durable agents—reliable LLM/tool execution, human control loops, and cross-boundary operations with Nexus.

Readers should be comfortable with distributed systems, microservice reliability, and production operations. The emphasis is on durable design, trade-offs, and operational reality: evolving workflow code safely through replay-first testing, debugging from hi

© 2026 NobleTrex Press (Rafbók): 6610001185374

Útgáfudagur

Rafbók: 13 mars 2026