Zabbix Kubernetes Discovery

Introduction

Kubernetes monitoring for Zabbix with discovery objects:

• Nodes
• DaemonSets
• Deployments
• StatefulSets
• Cronjobs
• PersistentVolumeClaims

Works with 2 variables only by default:

• ZABBIX_ENDPOINT: Zabbix server/proxy where the datas will be sent
• KUBERNETES_NAME: Name of your Kubernetes cluster on Zabbix (host)

Helm

Before installation, you need to create zabbix-monitoring namespace in your cluster:

$ kubectl create namespace zabbix-monitoring

All Helm options/parameters are available in the Helm folder here.

Install from local

To install the chart with the release name zabbix-kubernetes-discovery from local Helm templates:

$ helm upgrade --install zabbix-kubernetes-discovery \
    ./helm/zabbix-kubernetes-discovery/ \
    --values ./helm/zabbix-kubernetes-discovery/values.yaml \
    --namespace zabbix-monitoring \
    --set namespace.name="zabbix-monitoring" \
    --set environment.ZABBIX_ENDPOINT="zabbix-proxy.example.com" \
    --set environment.KUBERNETES_NAME="kubernetes-cluster-example"

Install from repo

To install the chart with the release name zabbix-kubernetes-discovery from my Helm repository:

$ helm repo add djerfy https://djerfy.github.io/helm-charts
$ helm upgrade --install zabbix-kubernetes-discovery \
    djerfy/zabbix-kubernetes-discovery \
    --namespace zabbix-monitoring
    --set namespace.name="zabbix-monitoring" \
    --set environment.ZABBIX_ENDPOINT="zabbix-proxy.example.com" \
    --set environment.KUBERNETES_NAME="kubernetes-cluster-name"

Uninstall

To uninstall/delete the zabbix-kubernetes-discovery deployment:

$ helm list -n zabbix-monitoring
$ helm delete -n zabbix-monitoring zabbix-kubernetes-discovery

The command removes all the Kubernetes components associated with the chart and deletes the release.

Commands

usage: zabbix-kubernetes-discovery.py [-h]
    [--zabbix-timeout ZABBIX_TIMEOUT]
    --zabbix-endpoint ZABBIX_ENDPOINT
    --kubernetes-name KUBERNETES_NAME
    --monitoring-mode {volume,deployment,daemonset,node,statefulset,cronjob}
    --monitoring-type {discovery,item,json}
    [--object-name OBJECT_NAME]
    [--match-label KEY=VALUE]
    [--exclude-name EXCLUDE_NAME]
    [--exclude-namespace EXCLUDE_NAMESPACE]
    [--no-wait]
    [--verbose]
    [--debug]

Zabbix

Import template

Zabbix template is located in ./zabbix/ folder on this repository.

After downloading, you need to import it as below:

1. Go to Configuration in menu
2. And Templates
3. Click Import
4. Select downloaded template file
5. Confirm import

Discovery rules

• Daemonset
  • Items: 4
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Available replicas
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Current replicas
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Desired replicas
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Ready replicas
  • Triggers: 5
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Available replicas nodata
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Current replicas nodata
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Desired replicas nodata
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Ready replicas nodata
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Problem items nodata
  • Graphs: 1
    • Daemonset {#KUBERNETES_DAEMONSET_NAME}: Graph replicas
• Deployment
  • Items: 3
    • Deployment {#KUBERNETES_DEPLOYMENT_NAME}: Available replicas
    • Deployment {#KUBERNETES_DEPLOYMENT_NAME}: Desired replicas
    • Deployment {#KUBERNETES_DEPLOYMENT_NAME}: Ready replicas
  • Triggers: 5
    • Deployment {#KUBERNETES_DEPLOYMENT_NAME}: Available replicas nodata
    • Deployment {#KUBERNETES_DEPLOYMENT_NAME}: Desired replicas nodata
    • Deployment {#KUBERNETES_DEPLOYMENT_NAME}: Ready replicas nodata
    • Deployment {#KUBERNETES_DEPLOYMENT_NAME}: Problem items nodata
    • Deployment {#KUBERNETES_DEPLOYMENT_NAME}: Problem number of replicas
  • Graphs: 1
    • Deployment {#KUBERNETES_DEPLOYMENT_NAME}: Graph replicas
• Statefulset
  • Items: 3
    • Statefulset {#KUBERNETES_STATEFULSET_NAME}: Available replicas
    • Statefulset {#KUBERNETES_STATEFULSET_NAME}: Desired replicas
    • Statefulset {#KUBERNETES_STATEFULSET_NAME}: Ready replicas
  • Triggers: 5
    • Statefulset {#KUBERNETES_STATEFULSET_NAME}: Available replicas nodata
    • Stetafulset {#KUBERNETES_STATEFULSET_NAME}: Desired replicas nodata
    • Statefulset {#KUBERNETES_STATEFULSET_NAME}: Ready replicas nodata
    • Statefulset {#KUBERNETES_STATEFULSET_NAME}: Problem items nodata
    • Statefulset {#KUBERNETES_STATEFULSET_NAME}: Problem number of replicas
  • Graphs: 1
    • Deployment {#KUBERNETES_STATEFULSET_NAME}: Graph replicas
• Cronjob
  • Items: 3
    • Cronjob {#KUBERNETES_CRONJOB_NAME}: Job exitcode
    • Cronjob {#KUBERNETES_CRONJOB_NAME}: Job restart
    • Cronjob {#KUBERNETES_CRONJOB_NAME}: Job reason
  • Triggers: 5
    • Cronjob {#KUBERNETES_CRONJOB_NAME}: Job exitcode nodata
    • Cronjob {#KUBERNETES_CRONJOB_NAME}: Job restart nodata
    • Cronjob {#KUBERNETES_CRONJOB_NAME}: Job reason nodata
    • Cronjob {#KUBERNETES_CRONJOB_NAME}: Problem items nodata
    • Cronjob {#KUBERNETES_CRONJOB_NAME}: Problem last job
  • Graphs: 1
    • Cronjob {#KUBERNETES_CRONJOB_NAME}: Graph jobs
• Node
  • Items: 8
    • Node {#KUBERNETES_NODE_NAME}: Allocatable cpu
    • Node {#KUBERNETES_NODE_NAME}: Allocatable memory
    • Node {#KUBERNETES_NODE_NAME}: Allocatable pods
    • Node {#KUBERNETES_NODE_NAME}: Capacity cpu
    • Node {#KUBERNETES_NODE_NAME}: Capacity memory
    • Node {#KUBERNETES_NODE_NAME}: Capacity pods
    • Node {#KUBERNETES_NODE_NAME}: Current pods
    • Node {#KUBERNETES_NODE_NAME}: Healthz
  • Triggers: 8
    • Node {#KUBERNETES_NODE_NAME}: Allocatable pods nodata
    • Node {#KUBERNETES_NODE_NAME}: Capacity pods nodata
    • Node {#KUBERNETES_NODE_NAME}: Current pods nodata
    • Node {#KUBERNETES_NODE_NAME}: Problem pods limits warning
    • Node {#KUBERNETES_NODE_NAME}: Problem pods limits critical
    • Node {#KUBERNETES_NODE_NAME}: Health nodata
    • Node {#KUBERNETES_NODE_NAME}: Health problem
    • Node {#KUBERNETES_NODE_NAME}: Problem items nodata
  • Graphs: 1
    • Node {#KUBERNETES_NODE_NAME}: Graph pods
• VolumeClaim
  • Items: 6
    • Volume {#KUBERNETES_PVC_NAME}: Available bytes
    • Volume {#KUBERNETES_PVC_NAME}: Capacity bytes
    • Volume {#KUBERNETES_PVC_NAME}: Capacity inodes
    • Volume {#KUBERNETES_PVC_NAME}: Free inodes
    • Volume {#KUBERNETES_PVC_NAME}: Used bytes
    • Volume {#KUBERNETES_PVC_NAME}: Used inodes
  • Triggers: 11
    • Volume {#KUBERNETES_PVC_NAME}: Available bytes nodata
    • Volume {#KUBERNETES_PVC_NAME}: Capacity bytes nodata
    • Volume {#KUBERNETES_PVC_NAME}: Capacity inodes nodata
    • Volume {#KUBERNETES_PVC_NAME}: Consumption bytes critical
    • Volume {#KUBERNETES_PVC_NAME}: Consumption bytes warning
    • Volume {#KUBERNETES_PVC_NAME}: Consumption inodes critical
    • Volume {#KUBERNETES_PVC_NAME}: Consumption inodes warning
    • Volume {#KUBERNETES_PVC_NAME}: Free inodes nodata
    • Volume {#KUBERNETES_PVC_NAME}: Used bytes nodata
    • Volume {#KUBERNETES_PVC_NAME}: Used inodes nodata
    • Volume {#KUBERNETES_PVC_NAME}: Problem items nodata
  • Graphs: 2
    • Volume {#KUBERNETES_PVC_NAME}: Graph bytes
    • Volume {#KUBERNETES_PVC_NAME}: Graph inodes

Development

Manual build

You can build Docker image manually like this:

$ docker build -t zabbix-kubernetes-discovery .

Contributing

All contributions are welcome! Please fork the main branch, create a new branch and then create a pull request.