Monitoring ZFS-LocalPV Volumes

Updated September 7th 2021:  This blog is updated with the the latest guide on OpenEBS Local PV, please refer to https://github.com/openebs/dynamic-localpv-provisioner.

Before reading this post, please read my previous post for instructions on setting up the ZFS-LocalPV for dynamically provisioning the volumes on the ZFS storage. Here, we will focus on how we can set up the Prometheus alert for Provisioned volumes when space utilization has reached a critical point.

Monitoring ZFS-LocalPV Volumes

Prerequisite

Make sure you are using k8s version 1.15+ to access the CSI volume metrics.

Setup helm

This step uses helm as the Kubernetes package manager. If you have not setup the helm, execute the below configuration. Otherwise, you can move on to the next step.

$ helm version
Client: &version.Version{SemVer:"v2.16.1", GitCommit:"bbdfe5e7803a12bbdf97e94cd847859890cf4050", GitTreeState:"clean"}
Server: &version.Version{SemVer:"v2.16.1", GitCommit:"bbdfe5e7803a12bbdf97e94cd847859890cf4050", GitTreeState:"clean"}

$ helm init
Tiller (the Helm server-side component) has been installed into your Kubernetes Cluster.

Please note: by default, Tiller is deployed with an insecure 'allow unauthenticated users' policy.
To prevent this, run `helm init` with the --tiller-tls-verify flag.
For more information on securing your installation see: https://docs.helm.sh/using_helm/#securing-your-helm-installation

$ kubectl create serviceaccount --namespace kube-system tiller
serviceaccount/tiller created

$ kubectl create clusterrolebinding tiller-cluster-rule --clusterrole=cluster-admin --serviceaccount=kube-system:tiller
clusterrolebinding.rbac.authorization.k8s.io/tiller-cluster-rule created

$ kubectl patch deploy --namespace kube-system tiller-deploy -p '{"spec":{"template":{"spec":{"serviceAccount":"tiller"}}}}'
deployment.extensions/tiller-deploy patched

Install Prometheus Operator

Once the helm is ready and the related tiller pods are up and running, use the Prometheus chart from the helm repository.

$ helm install stable/prometheus-operator --name prometheus-operator

Check all the required pods are up and running

$ kubectl get pods -l "release=prometheus-operator"
NAME                                                 READY   STATUS    RESTARTS   AGE
prometheus-operator-grafana-85bb5d49d-bffdg          2/2     Running   0          2m21s
prometheus-operator-operator-64844759f7-rpwws        2/2     Running   0          2m21s
prometheus-operator-prometheus-node-exporter-p9rl8   1/1     Running   0          2m21s

Setup alert rule

Check all the rules available in the system :-

$ kubectl get PrometheusRule
NAME                                                       AGE
prometheus-operator-alertmanager.rules                     4m21s
prometheus-operator-etcd                                   4m21s
prometheus-operator-general.rules                          4m21s
prometheus-operator-k8s.rules                              4m21s
prometheus-operator-kube-apiserver-error                   4m21s
prometheus-operator-kube-apiserver.rules                   4m21s
prometheus-operator-kube-prometheus-node-recording.rules   4m21s
prometheus-operator-kube-scheduler.rules                   4m21s
prometheus-operator-kubernetes-absent                      4m21s
prometheus-operator-kubernetes-apps                        4m21s
prometheus-operator-kubernetes-resources                   4m21s
prometheus-operator-kubernetes-storage                     4m21s
prometheus-operator-kubernetes-system                      4m21s
prometheus-operator-kubernetes-system-apiserver            4m21s
prometheus-operator-kubernetes-system-controller-manager   4m21s
prometheus-operator-kubernetes-system-kubelet              4m21s
prometheus-operator-kubernetes-system-scheduler            4m21s
prometheus-operator-node-exporter                          4m21s
prometheus-operator-node-exporter.rules                    4m21s
prometheus-operator-node-network                           4m21s
prometheus-operator-node-time                              4m21s
prometheus-operator-node.rules                             4m21s
prometheus-operator-prometheus                             4m21s
prometheus-operator-prometheus-operator                    4m21s

You can edit any of the default rules or create a new rule to get the alerts. Below is a sample rule to start generating alerts when available storage space is less than 10%.

apiVersion: monitoring.coreos.com/v1
kind: PrometheusRule
metadata:
  labels:
    app: prometheus-operator
    chart: prometheus-operator-8.5.4
    heritage: Tiller
    release: prometheus-operator
  name: prometheus-operator-zfs-alertmanager.rules
  namespace: default
spec:
  groups:
  - name: zfsalertmanager.rules
    rules:
    - alert: ZFSVolumeUsageCritical
      annotations:
        message: The PersistentVolume claimed by  in Namespace  is only % free.
      expr: |
        100 * kubelet_volume_stats_available_bytes{job="kubelet"}
          /
        kubelet_volume_stats_capacity_bytes{job="kubelet"}
           <10
      for: 1m
      labels:
        severity: critical

Now apply the above YAML so that Prometheus can fire the alerts when available space is less than 10%.

Check the Prometheus alert

To view the Prometheus web UI, you must expose it through a Service. A simple way to accomplish this is to use a Service of type NodePort.

$ cat prometheus-service.yaml
apiVersion: v1
kind: Service
metadata:
  name: prometheus-service
spec:
  type: NodePort
  ports:
  - name: web
    nodePort: 30090
    port: 9090
    protocol: TCP
    targetPort: web
  selector:
    prometheus: prometheus-operator-prometheus

Apply the above YAML

$ kubectl apply -f prometheus-service.yaml
service/prometheus-service created

Now you can access the alert manager UI via “node’s-external-ip:30090”.

$ kubectl get nodes -owide
NAME                                         STATUS   ROLES    AGE    VERSION          INTERNAL-IP   EXTERNAL-IP   OS-IMAGE             KERNEL-VERSION   CONTAINER-RUNTIME
gke-zfspv-pawan-default-pool-3e407350-xvzp   Ready    <none>   103m   v1.15.4-gke.22   10.168.0.45   34.94.3.140   Ubuntu 18.04.3 LTS   5.0.0-1022-gke   docker://19.3.2

Here, we can access the alert manager via URL: http://34.94.3.140:30090/

Check the Alert Manager

To view the Alert Manager web UI, expose it through a Service of type NodePort.

$ cat alertmanager-service.yaml
apiVersion: v1
kind: Service
metadata:
  name: alertmanager-service
spec:
  type: NodePort
  ports:
  - name: web
    nodePort: 30093
    port: 9093
    protocol: TCP
    targetPort: web
  selector:
    alertmanager: prometheus-operator-alertmanager

Apply the above YAML

$ kubectl apply -f alertmanager-service.yaml
service/alertmanager-service created

Now you can access the alert manager UI via “node’s-external-ip:30093”.

$ kubectl get nodes -owide
NAME                                         STATUS   ROLES    AGE    VERSION          INTERNAL-IP   EXTERNAL-IP   OS-IMAGE             KERNEL-VERSION   CONTAINER-RUNTIME
gke-zfspv-pawan-default-pool-3e407350-xvzp   Ready    <none>   103m   v1.15.4-gke.22   10.168.0.45   34.94.3.140   Ubuntu 18.04.3 LTS   5.0.0-1022-gke   docker://19.3.2

Again, we can access the alert manager via URL: http://34.94.3.140:30093/.

I hope you found this post to be useful. Feel free to contact me with any feedback or questions by using the comment section below.