You can simulate a failover for clusters with Test Failover and a regional outage for a multi-region cluster with Test Outage.
You can simulate a failover for clusters with Test Failover in the Atlas UI or API. On Atlas Core clusters, an election takes five seconds on average. On Atlas Infinite clusters, a standby node becomes the primary node without an election among the compute nodes. To learn more, see Test Primary Failover.
You can simulate a regional outage for a multi-region cluster with Simulate Regional Outage in the Atlas UI or API. The simulation runs for a configurable duration of 1, 3, or 7 days. To learn more, see Simulate Regional Outage.
You can simulate maintaining availability under sustained overload with the Load Shedding test in the Atlas UI. Load Shedding is part of Intelligent Workload Management, an Atlas availability system. The simulation runs for a configurable duration of 15 minutes, 1 hour, 8 hours, or 1 day. To learn more, see Test Cluster Load Shedding.
Required Access
You must have Project Cluster Resilience Tester,` Project Cluster Manager, or higher role to perform a Test Failover or Simulate Regional Outage.