Kubernetes zone redundancy

What this Scenario does

This Scenario drops network traffic between your Kubernetes nodes in one availability zone and the rest of your cluster, simulating a complete zone outage. This forces Kubernetes to reschedule pods and for your ingress controllers and service mesh to redirect traffic to pods running in the remaining zones.

‍

Why run this Scenario?

• Demonstrate that your Kubernetes cluster is truly zone-redundant, with pods distributed using topology spread constraints.
• Verify that Kubernetes ingress controllers and service meshes detect the zone outage and redirect traffic appropriately.
• Test whether pod anti-affinity rules ensure sufficient replicas in each zone to handle failover traffic.
• Validate that pod disruption budgets maintain service availability during the zone failure.

‍

Expected outcome

When a zone fails, Kubernetes automatically routes traffic to pods in secondary zones with no interruption in service.