A product manager at Temporal demonstrates how to set up multi-region failover for durable workflows using Temporal Cloud. The demo shows a simple TypeScript web page change detector running on Temporal in AWS US-East-1, adding a US-East-2 replica namespace, simulating an outage by killing all US-East-1 workers, triggering a namespace failover in the Temporal UI, and cold-starting a worker in US-East-2 to resume workflows with full event history intact. The entire failover process takes about 3 minutes of downtime, and failback is equally straightforward.
•6m watch time
Sort: