top of page
Writer's pictureSathish Kumar

Kubernetes Administration - DevOps- Monitoring with Grafana


Note: If you have missed my previous articles on Docker and Kubernetes, you can find them here:  
Application deployment models evolution.
Getting started with Docker.Docker file and images.
Publishing images to Docker Hub and re-using them.
Docker- Find out what's going on.
Docker Networking- Part 1.
Docker Networking- Part 2.
Docker Swarm-Multi-Host container Cluster.Docker Networking- Part 3 (Overlay Driver).
Introduction to Kubernetes.Kubernetes- Diving in (Part 1)-Installing Kubernetes multi-node cluster.
Kubernetes-Diving in (Part2)- Services.
Kubernetes- Infrastructure As Code with Yaml (part 1).
Kubernetes- Infrastructure As Code Part 2- Creating PODs with YAML.
Kubernetes Infrastructure-as-Code part 3- Replicasets with YAML.
Kubernetes Infrastructure-as-Code part 4 - Deployments and Services with YAML.
Deploying a microservices APP with Kubernetes.
Kubernetes- Time based scaling of deployments with python client.
Kubernetes Networking - The Flannel network explained.
Kubernetes- Installing and using kubectl top for monitoring nodes and PoDs
Kubernetes Administration- Scheduling
Kubernetes Administration- Storage
Kubernetes Administration- Users
Kubernetes Administration - Network Policies with Calico network plugin
Kubernetes Administration - Managing Kubernetes Clusters with Rancher 
Kubernetes Administration - Package Management with Helm
Kubernetes Administration - Monitoring cluster health with Prometheus 
Kubernetes Administration - DevOps- Prometheus- The need and high-level architecture

Prometheus is a powerful tool that can be used to monitor nodes and microservice applications. Prometheus can be installed with Helm and its architecture enables third-party tools supporting PromQL to query data. Grafana is an observability tool that can be used to visualize Prometheus data. In addition to supporting Prometheus, Grafana also supports datasets from various other sources like cloud providers, hosts, etc.


If you intended to follow along, ensure that you have Helm installed- you can refer to my previous article for this.


Installing Grafana


The simplest way to install Grafana is with a helm chart for Kube-Prometheus-stack (Grafana is part of the install). This chart installs Prometheus, Grafana, and a bunch of other tools that enable easy monitoring.



root@sathish-vm2:/home/sathish# helm repo add prometheus-community https://prometheus-community.github.io/helm-charts 

root@sathish-vm2:/home/sathish# helm repo update
Hang tight while we grab the latest from your chart repositories...
...Successfully got an update from the "prometheus-community" chart repository
...Successfully got an update from the "stable" chart repository
Update Complete. ⎈Happy Helming!⎈

root@sathish-vm2:/home/sathish# helm install prometheous-vm prometheus-community/kube-prometheus-stack
NAME: prometheous-vm
LAST DEPLOYED: Sun Apr  4 07:49:44 2021
NAMESPACE: default
STATUS: deployed
REVISION: 1
NOTES:
kube-prometheus-stack has been installed. Check its status by running:
  kubectl --namespace default get pods -l "release=prometheous-vm"

Visit https://github.com/prometheus-operator/kube-prometheus for instructions on how to create & configure Alertmanager and Prometheus instances using the Operator.

root@sathish-vm2:/home/sathish/example-voting-app# kubectl get pods
NAME                                                     READY   STATUS    RESTARTS   AGE
alertmanager-prometheous-vm-kube-promet-alertmanager-0   2/2     Running   0          16s
prometheous-vm-grafana-75cdd4c75-qv9wd                   2/2     Running   0          19s
prometheous-vm-kube-promet-operator-55d66df7db-bll85     1/1     Running   0          19s
prometheous-vm-kube-state-metrics-858f884bc8-dxknj       1/1     Running   0          19s
prometheous-vm-prometheus-node-exporter-bmz5p            1/1     Running   0          19s
prometheous-vm-prometheus-node-exporter-lnnhq            1/1     Running   0          19s
prometheus-prometheous-vm-kube-promet-prometheus-0       2/2     Running   1          15s

Accessing the service


All the installed services are of type ClusterIP and cannot be accessed outside of the cluster. If you are running it on the cloud, it is best to use Ingress to expose Grafana and Prometheus. For local networks, you can use the NodePort service type.


root@sathish-vm2:/home/sathish/example-voting-app# kubectl get  svc
NAME                                                 TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                      AGE
alertmanager-operated                                ClusterIP   None             <none>        9093/TCP,9094/TCP,9094/UDP   61s
kubernetes                                           ClusterIP   10.96.0.1        <none>        443/TCP                      171d
prometheous-vm-grafana                               ClusterIP   10.105.48.224    <none>        80/TCP                       65s
prometheous-vm-kube-promet-alertmanager              ClusterIP   10.107.67.154    <none>        9093/TCP                     64s
prometheous-vm-kube-promet-operator                  ClusterIP   10.102.177.200   <none>        443/TCP                      64s
prometheous-vm-kube-promet-prometheus                ClusterIP   10.106.158.240   <none>        9090/TCP                     64s
prometheous-vm-kube-state-metrics                    ClusterIP   10.96.39.244     <none>        8080/TCP                     64s
prometheous-vm-prometheus-node-exporter              ClusterIP   10.104.169.34    <none>        9100/TCP                     64s
prometheus-operated                                  ClusterIP   None             <none>        9090/TCP                     60s
prometheus-prometheous-vm-kube-promet-prometheus-0   ClusterIP   10.102.92.28     <none>        9090/TCP,8080/TCP            2m29s


Adding a Data Source


As I mentioned before, Grafanna can collect data from various data sources. What we are interested in is Prometheus. Click the "Add your first data source" option and add Prometheus.


My Prometheus service is of type ClusterIP, so I need to use the corresponding IP


root@sathish-vm2:/home/sathish# kubectl get svc
NAME                                      TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                       AGE
alertmanager-operated                     ClusterIP   None             <none>        9093/TCP,9094/TCP,9094/UDP    22m
kubernetes                                ClusterIP   10.96.0.1        <none>        443/TCP                       171d
prometheous-vm-grafana                    NodePort    10.96.123.173    <none>        80:31389/TCP,3000:30067/TCP   15m
prometheous-vm-kube-promet-alertmanager   ClusterIP   10.104.94.128    <none>        9093/TCP                      22m
prometheous-vm-kube-promet-operator       ClusterIP   10.101.130.227   <none>        443/TCP                       22m
prometheous-vm-kube-promet-prometheus     ClusterIP   10.102.234.84    <none>        9090/TCP                      22m
prometheous-vm-kube-state-metrics         ClusterIP   10.104.106.132   <none>        8080/TCP                      22m
prometheous-vm-prometheus-node-exporter   ClusterIP   10.100.242.56    <none>        9100/TCP                      22m
prometheus-operated                       ClusterIP   None             <none>        9090/TCP                      22m

Note: It might be a good idea to change the service type of Prometheus UI, so it can be accessed from another machine. I have changed it to NodePort on my setup



root@sathish-vm2:/home/sathish/example-voting-app# kubectl  get svc
NAME                                                 TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                         AGE
alertmanager-operated                                ClusterIP   None             <none>        9093/TCP,9094/TCP,9094/UDP      15m
kubernetes                                           ClusterIP   10.96.0.1        <none>        443/TCP                         171d
prometheous-vm-grafana                               NodePort    10.104.141.104   <none>        80:30067/TCP                    12m
prometheous-vm-kube-promet-alertmanager              ClusterIP   10.102.236.148   <none>        9093/TCP                        15m
prometheous-vm-kube-promet-operator                  ClusterIP   10.102.57.69     <none>        443/TCP                         15m
prometheous-vm-kube-promet-prometheus                ClusterIP   10.100.201.230   <none>        9090/TCP                        15m
prometheous-vm-kube-state-metrics                    ClusterIP   10.105.155.65    <none>        8080/TCP                        15m
prometheous-vm-prometheus-node-exporter              ClusterIP   10.96.183.206    <none>        9100/TCP                        15m
prometheus-operated                                  ClusterIP   None             <none>        9090/TCP                        15m
prometheus-prometheous-vm-kube-promet-prometheus-0   NodePort    10.102.92.28     <none>        9090:30090/TCP,8080:31151/TCP   2m24s


Once the data source is added, you can create a dashboard to monitor various things.



Prometheus monitors a list of objects exposed by the Service Monitor object. Here is a list of default service monitor objects that the Helm chart installs.


root@sathish-vm2:/home/sathish/example-voting-app# kubectl get servicemonitor
NAME                                                 AGE
prometheous-vm-kube-promet-alertmanager              19m
prometheous-vm-kube-promet-apiserver                 19m
prometheous-vm-kube-promet-coredns                   19m
prometheous-vm-kube-promet-grafana                   19m
prometheous-vm-kube-promet-kube-controller-manager   19m
prometheous-vm-kube-promet-kube-etcd                 19m
prometheous-vm-kube-promet-kube-proxy                19m
prometheous-vm-kube-promet-kube-scheduler            19m
prometheous-vm-kube-promet-kube-state-metrics        19m
prometheous-vm-kube-promet-kubelet                   19m
prometheous-vm-kube-promet-node-exporter             19m
prometheous-vm-kube-promet-operator                  19m
prometheous-vm-kube-promet-prometheus                19m

The release label registers the serviceMonitor with Prometheus enabling it to collect data from a particular serviceMonitor object.

root@sathish-vm2:/home/sathish/example-voting-app# kubectl describe servicemonitor prometheous-vm-kube-promet-grafana
Name:         prometheous-vm-kube-promet-grafana
Namespace:    default
Labels:       app=kube-prometheus-stack-grafana
              app.kubernetes.io/managed-by=Helm
              chart=kube-prometheus-stack-14.5.0
              heritage=Helm
              release=prometheous-vm

Monitoring Apps


Grafanna can be used to monitor the status of various components of applications. I have already deployed the voting app (described here ) and now components of app can be monitored with Grafanna. To enable the app component to export data an exporter is required. An exporter converts data to a format that Prometheus understands and also exposes a metrics endpoint. A list of prebuilt exporters is available here:



If desirable exporter for a particular app component can be installed after customizing parameters.

helm show values prometheus-community/prometheus-postgres-exporter > postgres.yaml

# Edit and change the highlighted fields

# Default values for prometheus-postgres-exporter.
# This is a YAML-formatted file.
# Declare variables to be passed into your templates.

replicaCount: 1

image:
  repository: quay.io/prometheuscommunity/postgres-exporter
  tag: v0.9.0
  pullPolicy: IfNotPresent

  ## Optionally specify an array of imagePullSecrets.
  ## Secrets must be manually created in the namespace.
  ## ref: https://kubernetes.io/docs/tasks/configure-pod-container/pull-image-private-registry/
  ##
  # pullSecrets:
  #   - myRegistrKeySecretName

service:
  type: ClusterIP
  port: 80
  targetPort: 9187
  name: http
  labels:
    release: prometheous-vm
  annotations: {}
...........................................
  datasource:
    # Specify one of both datasource or datasourceSecret
    #ClusterIP of DB pod
    host: 10.101.175.235
    user: postgres
    # Only one of password and passwordSecret can be specified
    password: postgres

#Deploy the chart

root@sathish-vm2:~# helm install postgres-exporter prometheus-community/prometheus-postgres-exporter -n vote -f postgres.yaml
NAME: postgres-exporter
LAST DEPLOYED: Sun Apr  4 11:12:44 2021
NAMESPACE: vote
STATUS: deployed
REVISION: 1
TEST SUITE: None
NOTES:
1. Get the application URL by running these commands:
  export POD_NAME=$(kubectl get pods --namespace vote -l "app=prometheus-postgres-exporter,release=postgres-exporter" -o jsonpath="{.items[0].metadata.name}")
  echo "Visit http://127.0.0.1:8080 to use your application"
  kubectl port-forward $POD_NAME 8080:80

Grafana dashboards


Grafana has a list of prebuilt dashboards that can be used to monitor the status of various resources.





Customized dashboards can also be created from UI.



Playlists could be created from UI that automatically circles through dashboards- this could be useful for DevOps operators.




Grafanna is a very useful tool for DevOps monitoring and is widely deployed. I addition to the capabilities described in this article, Grafanna supports lots of features and plugins. More details about Grafanna can be found here.


Hope this short introduction to Grafana was useful. Till next time, ciao and have a great week ahead.


295 views0 comments

Komentáre


bottom of page