Skip to content
A set of dashboards and panels for kubernetes.
Branch: master
Clone or download
daniellee Spelling on config page
Didn't update dist - it can be included next time the plugin is released.
Latest commit 6424c15 Dec 17, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
dist remove legacy slugify functions Mar 23, 2018
src Spelling on config page Dec 17, 2018
.gitattributes Don't diff files in dist/ dirrectory. Sep 30, 2016
.gitignore convert to typescript Jan 26, 2018
Gruntfile.js
LICENSE
README.md
package.json add template var capability to kubernetes plugin Jan 31, 2018
tsconfig.json convert to typescript Jan 26, 2018
tslint.json
yarn.lock add template var capability to kubernetes plugin Jan 31, 2018

README.md

Grafana App for Kubernetes

Kubernetes is an open-source system for automating deployment, scaling, and management of containerized applications.

The Grafana Kubernetes App allows you to monitor your Kubernetes cluster's performance. It includes 4 dashboards, Cluster, Node, Pod/Container and Deployment. It allows for the automatic deployment of the required Prometheus exporters and a default scrape config to use with your in cluster Prometheus deployment. The metrics collected are high-level cluster and node stats as well as lower level pod and container stats. Use the high-level metrics to alert on and the low-level metrics to troubleshoot.

Container Dashboard

Container Dashboard

Node Dashboard

Requirements

  1. Currently only has support for Prometheus
  2. For automatic deployment of the exporters, then Kubernetes 1.6 or higher is required.
  3. Grafana 5.0.0+

Features

  • The app uses Kubernetes tags to allow you to filter pod metrics. Kubernetes clusters tend to have a lot of pods and a lot of pod metrics. The Pod/Container dashboard leverages the pod tags so you can easily find the relevant pod or pods.

  • Easy installation of exporters, either a one click deploy from Grafana or detailed instructions to deploy them manually them with kubectl (also quite easy!)

  • Cluster level metrics that are not available in Heapster, like CPU Capacity vs CPU Usage.

Cluster Metrics

  • Pod Capacity/Usage
  • Memory Capacity/Usage
  • CPU Capacity/Usage
  • Disk Capacity/Usage
  • Overview of Nodes, Pods and Containers

Node Metrics

  • CPU
  • Memory Available
  • Load per CPU
  • Read IOPS
  • Write IOPS
  • %Util
  • Network Traffic/second
  • Network Packets/second
  • Network Errors/second

Pod/Container Metrics

  • Memory Usage
  • Network Traffic
  • CPU Usage
  • Read IOPS
  • Write IOPS

Documentation

Installation

  1. Use the grafana-cli tool to install kubernetes from the commandline:
grafana-cli plugins install grafana-kubernetes-app
  1. Restart your Grafana server.

  2. Log into your Grafana instance. Navigate to the Plugins section, found in the Grafana main menu. Click the Apps tabs in the Plugins section and select the newly installed Kubernetes app. To enable the app, click the Config tab and click on the Enable button.

Connecting to your Cluster

  1. Go to the Cluster List page via the Kubernetes app menu.

    Cluster List in main menu

  2. Click the New Cluster button.

  3. Fill in the Auth details for your cluster.

    TLS certs/keys must be provided in plaintext, not base64 encoded form. For example:

    -----BEGIN CERTIFICATE-----
    MIQWQtAEFeqqfAFeAEGEQWIGNwEQNFGQ4AEFN35AKWadgAENGqiEGNIWm1QETDGF
    ...
    -----END CERTIFICATE-----
    
  4. Choose the Prometheus datasource that will be used for reading data in the dashboards.

  5. Click Deploy. This will deploy a Node Exporter DaemonSet, to collect health metrics for every node, and a Deployment that collects cluster metrics.

Feedback and Questions

Please submit any issues with the app on Github.

You can’t perform that action at this time.