How to build distributed systems in JavaScript
summary
Shows a working order-processing example that replaces a serverless function plus MongoDB state machine, retries and compensations with a Temporal workflow and small activity functions. Demonstrates workflow id for idempotency, automatic infinite retry exponential backoff for activities, durable state persisted via event sourcing, and durable 30‑day timers so long‑running billing survives restarts. Claims fewer incidents and simpler code; the demo runs: client starts a workflow, workers execute activities, the Temporal cluster resumes failed runs and guarantees completion.
AI-written summary · about
chapters (16)
- 0:00Introduction to talk
- 1:17What is a distributed system
- 4:26Reliability challenges in microservices
- 8:40Adding retries and timeouts
- 11:06Handling compensation steps
- 15:21State persistence and recovery
- 16:25Full manual implementation complexity
- 17:07Simplifying with Temporal
- 20:52Why use Temporal
- 24:48Orchestration vs choreography
- 27:14Programming model with code
- 31:26Long-running workflows and timers
- 34:15Temporal APIs and tools
- 35:17Benefits of switching to Temporal
- 37:45When to use Temporal
- 40:15Q&A
related talks
01h 38mBuilding Reliable Distributed Systems
26mMicroservices Madness: Application And Infrastructure Automation Using Temporal
30mOrchestrate Your Business Workflows with Temporal
23mDesigning a Workflow Engine From First Principles
50mThe end of event-based architectures? Orchestration with Temporal
06mDurable Executions in the Face of (some) Failures
32mWhy idempotency matters during durable execution in modern software stacks
26mTemporal: A Better Way to Build Modern Applications