admin:devops:start

Infrastructure as code / DevOps

Every change to a system should be done programmatically and saved in a code repository, so you can replay it when needed. You then also know what happened on your servers because there are no manual tasks done to them.

There should be an automated deployment where you can quickly recover from disasters with backups and infrastructure as code, and doing lots of time intensive manual tasks on individual systems should be replaced by automated recovery.1

  • Puppet
    • Bolt – agentless puppet
  • Ansible – Written in Python, by Red Hat
  • Chef Infra – written in Ruby
  • Saltstack
  • Last modified: 2022-06-29 14:29