OpenShift Infra Nodes, Taints, and Tolerations

Here is how you set up worker nodes to be infra nodes and run infrastructure workloads in OpenShift 4.x

Sections:

0:00 - Start
0:10 - Labeling Nodes with Infra
2:55 - Setting Affinity of Ingress Controller to Infra Nodes
4:57 - Tainting the Infra Nodes to Reserve them for OCP Workloads
5:54 - Forcing Pods to Redeploy on App Nodes
6:53 - Now fix the Ingress Workload to Tolerate the NoSchedule Taint
9:10 - Fix the cacheing Image Registry to Affinitize the Infra Node
11:00 - Fix the Registry with a Toleration for the Infra Node
11:50 - Affinitize the Monitoring Workloads to Infra Nodes
14:32 - Fix the Monitoring Config Map to Provide a Toleration for Infra Nodes