Almost a decade ago I started working on a side project to transition my wordpress fleet to a single server side rendering microservice I call the multiDomainCMS. Having 100+ domains in my portfolio meant I needed automation if I would be successful at managing a fleet. Without my virtual assistants things fell apart. And I felt like there was a better way to architect everything. First wp_settings and wp_posts migration began into mongo collections. Then I began to transition themes over to a nodeSass system (now deprecated) and template code into ReactJS. I wired ExpressJS to partition logic and route traffic to content and theme based on slug and domain. That’s it, that was the core idea. I’ve been fleshing it out with all kinds of other tools. And am even considering a sharding strategy similar to Slack, as the solution for a turnkey webmaster system. It’s working for me really well. Rapid development with Antigravity and a trunk based workflow. No more Jenkins, custom job-runner for zero-downtime deployments via github webhook. Custom IDM and Orders microservices, business continuity and DR. I have Grafana and Prometheus set up for observability and was surprised with insights I gained about the internet as a whole, as a benefit from all my logs going through a central dashboard. I’ve been making youtube videos about it. And my agent told me I should post here. Honestly I am scared to death of sounding stupid. But after we did some benchmarks, it said this community would want to hear about it. I pressed it real hard on collecting real scientific metrics in a head to head stress test against a wordpress version of one of my sites. Feel free to poke holes, make suggestions, ask questions about the benchmark data or anything else. Thanks in advance.
1 comments