Skip to content

2024

Implementing the Blog Plugin

I had decided that I wanted to experiment with the built-in blog functionality of Material MKDocs because it might be a way to catalog my thoughts as I work on projects. A lot of time, my notes are scattered across RocketChat and Discord, and get thrown into disarray, so I hope that I can instead centralize my thoughts, projects, and things of that sort here. This is the first step on that journey.

The blog workflow

When I write a blog article, it is generally based on the official Blog Documentation. The formatting uses a few metadata tags, and looks like the following:

---
draft: false
date: 2024-12-15
updated: 2024-12-15
authors:
  - nicole
categories:
  - General
  - Documentation
tags:
  - MKDocs
  - Material MkDocs
  - Documentation
---

# Example Post Title
Placeholder Text / Body of the Blog Post.

Openstack Frustrations

So, I want to start with a little context. As part of a long-standing project I have been working on, I have tried to deploy OpenStack. OpenStack is sort of envisioned as "Infrastructure as a Service (IAAS)". Basically you deploy an OpenStack cluster, which can run its own KVM for virtual machine and containers, or it can interface with an existing Hypervisor infrastructure, such as Hyper-V. In most cases, people branch out the "Control", "Compute", and "Storage" roles into different physical servers, but in my homelab, I have been attempting to deploy it via a "Converged" model, of having Control, Compute, and Storage on each node, spanning a high-availability cluster of 3 nodes.

The Problem

The problems come into the overall documentation provided for deploying either Canonical Openstack which I have detailed my frustrations of the system in my own attempted re-write of the documentation here. I have also attempted to deploy it via Ansible OpenStack, whereas my documentation thus far in my homelab is visible here.

You see, OpenStack is like icecream, it has many different ways to deploy it, and it can be as simple, or as overtly-complex as you need it to be, and it scales really well across a fleet of servers in a datacenter. My problems come in where the Canonical deployment has never worked fully / properly, and their own development team is hesitant to recommend the current documentation, and the Ansible OpenStack deployment process, while relatively simple, requires a base of existing knowledge that makes translating the instructions into more user-friendly instructions in my homelab documentation a difficult task. Eventually I want to automate much of the process as much as I can, but that will take time.

The common issue I've seen while trying to deploy OpenStack is understanding the networking, how networking is configured, network bridges, etc. The process is different based on the deployment method (Currently trying to deploy it via OpenStack Ansible). Hopefully in the near future I will make some kind of breakthrough in the deployment process and get everything working.

I will post an update later if I figure things out!