Skip to content
forked from redhat-cop/osia

Tool for reliable automated deployments of OpenShift Container Platform 4.x into OpenStack and AWS.

License

Notifications You must be signed in to change notification settings

lapd-devops/osia

 
 

Repository files navigation

OSIA

OpenShift infra automation.

Goal

The tool aims to create unified installer

Main features

  • Find empty region in aws to install opneshift on
  • Find feasible network in PSI and allocate FIPs before installation happens
  • Generate install-config.yaml from predefined defaults.
  • Store generated files for deletion to git repository and push changes right after the cluster is installed
  • Manage DNS entries based on the installation properties and results
  • Clean everything once the cluster is not needed

Usage

The tool operates over directory which is expected to be git repository and where the service will store generated configuration and push it to the upstream repository of currently working branch.

Common configuration

The common configuraiton is done by yaml file called settings.yaml that should be located at CWD (root of the repository in most cases).

The configuration has following structure:

default:
  cloud:
    openstack:
      base_domain: ''
      certificate_bundle_file: ''
      pull_secret_file: ''
      ssh_key_file: ''
      osp_cloud: ''
      osp_base_flavor: ''
      network_list: []
    aws:
      base_domain: ''
      pull_secret_file: ''
      certificate_bundle_file: ''
      ssh_key_file: ''
      worker_flavor: '' 
      list_of_regions: []
  dns:
    route53:
      ttl: 0
    nsupdate:
      server: ''
      zone: ''
      key_file: ''
      ttl: 0 

Every key here is overridible by the argument passed to the installer. For explanation of any key, please check he documentation below.

About

Tool for reliable automated deployments of OpenShift Container Platform 4.x into OpenStack and AWS.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 93.8%
  • HTML 6.2%