I am deploying rstudio connect on kubernetes and unable to add tolerations from the helm chart. When I check
kubectl describe pod rstudio-connect
The following tolerations are shown.
Tolerations: node.kubernetes.io/not-ready:NoExecute op=Exists for 300s
node.kubernetes.io/unreachable:NoExecute op=Exists for 300s
Here is how I am applying the tolerations and affinity from the helm chart.
pod:
tolerations:
- key: "workload"
operator: "Equal"
value: "cpu-compute"
effect: "NoSchedule"
affinity:
nodeAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
nodeSelectorTerms:
- matchExpressions:
- key: intent
operator: In
values:
- cpu-compute
How can I apply the tolerations from the helm chart? I think the affinity is getting applied because the pod becomes unschedulable. The reason it cannot provision a node is because the affinity is cpu-compute but the toleration is not.