Processes, StateCharts and WorkFlows
summary
A concrete recipe for long‑running work shows a React storefront that starts as one endpoint, breaks into many, and then needs durable coordination: implement a saga with compensating transactions by wrapping each step with its compensator, persist step ids to make retries idempotent, and use messaging either by orchestration vs choreography. Shift from transactions to explicit states using StateCharts as JSON DSL, demonstrate a step function that pushes compensations on a stack, and use sharding to avoid duplicate effect execution.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:11Italian restaurant story
- 4:04Long-running processes
- 8:19Human workflows
- 9:01ACID transactions fail
- 13:14Saga pattern
- 16:49Orchestration vs choreography
- 17:31Failures and compensations
- 21:03Compensating transactions
- 23:13Persistence and re-execution
- 26:29StateCharts and inconsistency
- 28:56Evolving processes
- 31:47Side effects and concurrency
- 33:13Recap
related talks
46mLost in transaction? Strategies to deal with (in-)consistency in distributed systems
58mWelcome to the (state) machine
25mReactor Under the Hood: Building a Graph-Based Saga Orchestrator in Elixir
45mRedux and Redux-saga to manage states in a large React app
27mReact, But For Data
51mBe proactive about state in a reactive system
27mComplexity: "Divide and Conquer!"
32mStructuring Your App’s Story With Sagas and Selectors