Hi again!

It’s been a little while. So what’s this all about, you ask?

I’ve decided to move the blog to a headless system, notably Gatsby. The move was spurred on by my own personal requirement to decouple my web application, mitigate security risks, lower costs and drive continual improvement quickly and easily.

What was done?

I’ve now moved the blog to a headless system powered by Gatsby. It’s a react-based framework that will generate static pages from react elements. These static pages are then hosted on the provider of your choice! (S3, Amplify, Netlify, etc.). The real power that comes from this is there are no servers to manage, update, protect and pay for.

The blog is now hosted via AWS Amplify and distributed globally with AWS CloudFront, providing high-speed access and low administration. In addition, the pages are static in themselves and can be updated by pushing a new update to a Git repo. The Git repo triggers an AWS Amplify build and rebuilds the project, generating new static pages and invalidating any CloudFront caches where required.

Why?

Cost mostly. I was previously using a T3 Micro EC2 instance to host the previous version, which would rack up an easy $15+ a month bill.

The instance would host the WordPress application, the Apache server and SQL DB. So having a single point of failure is never a good thing. Other options came to mind concerning v2. One was using AWS Fargate, hosting the containers serverless’ly and hosting the DB on AWS RDS. This, in theory, would lower costs again, but still a bit too complex for a simple blog.

Another option was to use Gatsby and use the S3 bucket as the destination for the static project files, but this could require updating all items in the bucket every time a build is needed and could get fiddly.

So, I settled on Amplify. This would take away the job of building the application, provide CI & lower costs significantly. Approximate prices with the new setup should be less than $0.90.

If I wanted to test new features on the blog, I could also easily create a feature branch, and test without affecting the production environment, and merge if stable. That’s something for another day.

Anything else going on?

I recently passed my AWS CSAA!

After a two-month study crunch, I managed to pass the exam. I initially studied for this exam about two years ago, but Covid had other plans regarding my exam. Nevertheless, I’m delighted with the accomplishment and am now looking toward the AWS Advanced Network Speciality exam.