This is a blog with random OpenShift, Kubernetes, OpenStack and Linux related notes so I don't forget things. If you find something inaccurate or that could be fixed, please file a bug report here.

View on GitHub

Back to home

27 August 2018

TripleO service deployment steps

by Juan Antonio Osorio Robles

TripleO does a lot of things in the whole deployment process, from setting up the appropriate networks, to provisioning the baremetal hosts. However, there is a significant part of the deployment in which services are configured. This part of the deployment is done in 5 steps. What happens in these steps has traditionally been the following:

1) Load Balancer configuration

2) Core Services (Database/Rabbit/NTP/etc.)

3) Early Openstack Service setup (Ringbuilder, etc.)

4) General OpenStack Services

5) Service activation (Pacemaker) & Service post-configuration (e.g. nova host discovery)

This has tended to be done by Puppet, however, with the arrival of containers to the deployment, and the need to provide more options to deploy the services, there are now several things happening in between these steps. These would be the overall steps nowadays:

It’s useful to have the overall picture. Next, I’ll try to explain the details that are relevant to understanding what happens in each step and before them.

With these brief explanations, I hope you have a better notion of what’s going on in the deployment steps. If you want more in-depth knowledge about these steps, I’m also writing a services of blog posts that describe the service templates; which include way more detail. The series starts with a blog post describing the minimal things you need for your service template

tags: tripleo - openstack

Back to home