kubeojo
kubeojo copied to clipboard
Health Analytics for testsuites with elixir and d3js (inspired by CHAOSS)
"What is kubeojo?" in 10 seconds
Developing kubeojo
See devel-setup
How to run kubeojo
At this point kubeojo is under active development and has not yet reached version 1.0. As soon as a stable version is ready, the docs will be updated to describe how to run kubeojo. ( Please feel free to pick up issues from GitHub if you want to contribute for version 1.0)
What is kubeojo?
Nowadays, it is extremely common that projects have a test suite running on CI system to ensure the project's stability.
When test suites get big, it is very often the case that there are "brittle tests" in the suite. These are the kinds of tests that typically run just fine, but randomly fail from time to time (such as once every 10 runs or so).
Kubeojo methodically examines your tests failures. It stores the results and visualizes the "brittle tests" to help detect and fix them. The UI of kubeojo is powered by Phoenix 1.3 and D3.js.
Configuration:
In order to use kubeojo, you need to have 2 YAML files configured:
- Jenkins credentials:
kubeojo/kubeojo/config/jenkins_credentials.yml
as password you can use the a Jenkins Token:
jenkins_url: "https://i_love_opensuse.ci.com/"
username: "Jenkins_username"
password: "2faidfakjfdkjadf30ff"
- Jenkins Jobs you want to analyze:
kubeojo/kubeojo/config/jenkins_jobs.yml
Here you should insert the name of the jobs you want to analyze the tests results for.
Important: your jobs need to export tests in junit-format to ensure kubeojo can fetch the junit_results.
jenkins_jobs: ["manager-3.1-cucumber", "manager-Head-cucumber"]
Roadmap:
https://github.com/MalloZup/kubeojo/issues
Contributors:
Thanks to all contributors for kubeojo!