Recommended Scenario
Kubernetes memory scalability
Verify that your Kubernetes service scales as expected when memory is limited. Gremlin increases memory utilization in three stages—50%, 75%, and 90%—to validate Kubernetes memory management.
Experiment types
Memory
Targets
Kubernetes
Length
20 minutes
How it works
How this Scenario works
This Scenario progressively consumes memory within your Kubernetes pods in three stages: 50%, 75%, and 90%. This simulates memory pressure from traffic spikes, memory leaks, or under-provisioned pod resource limits.
Use cases
Why run this Scenario?
- Identify memory leaks or poor garbage collection within pod workloads before they trigger OOMKiller.
- Test how Kubernetes memory resource limits and requests interact with the OOMKiller under pressure.
- Verify that Kubernetes pod autoscaling (HPA or VPA) responds to memory pressure as configured.
- Optimize pod memory resource limits to balance cost, performance, and cluster stability.
Result
What to expect when you run it
If memory usage spikes to 90% within a Kubernetes pod, the application continues to function without performance degradation for at least 5 minutes.
Keep going
More Scenarios to try
Avoid downtime. Use Gremlin to turn failure into resilience.
Gremlin empowers you to proactively root out failure before it causes downtime. See how you can harness chaos to build resilient systems by requesting a demo of Gremlin.
.svg)