Back to all scenarios
Scenario #168
Networking
K8s v1.22, On-premise

Cluster Network Downtime Due to CNI Plugin Upgrade

Cluster network downtime occurred during a CNI plugin upgrade, affecting pod-to-pod communication.

Find this helpful?
What Happened

During an upgrade to the CNI plugin, the network was temporarily disrupted due to incorrect version compatibility and missing network configurations.

Diagnosis Steps
  • 1Inspected pod logs and noticed failed network interfaces after the upgrade.
  • 2Checked CNI plugin version compatibility and identified missing configurations for the new version.
Root Cause

The new version of the CNI plugin required additional configuration settings that were not applied during the upgrade.

Fix/Workaround
• Applied the required configuration changes for the new CNI plugin version.
• Restarted affected pods and network components to restore connectivity.
Lessons Learned

Always verify compatibility and required configurations before upgrading the CNI plugin.

How to Avoid
  • 1Test plugin upgrades in a staging environment to catch compatibility issues.
  • 2Follow a defined upgrade process that includes validation of configurations.