How to update the resource limits/requests in a persistent way.

We need to edit the tectonic-monitoring ConfigMap. Here's the relevant doc: Configuring Tectonic Monitoring

Let's watch the yaml output of the prometheus/k8s:

watch -n 3 kubectl -n tectonic-system get prometheus/k8s -o yaml

Then in another terminal use kubectl to edit the configmap:

kubectl -n tectonic-system edit cm/tectonic-monitoring

Under the prometheusK8s section add a resources definition. Be sure to include both cpu and memory limits and requests. Here is an example that doubles the memory values:

    prometheusK8s:
      baseImage: quay.io/prometheus/prometheus
      resources:
        limits:
          cpu: 400m
          memory: 4000mi
        requests:
          cpu: 100m
          memory: 1000mi

The prometheus/k8s should update automatically. No need to recycle the pod.