I'm currently working at a company that builds WordPress websites. We've been using a traditional setup with EC2, and the devs are still updating directly via FTP, but I'm focusing on improving our infrastructure. I recently switched our architecture to use containers on ECS Fargate, which involved creating a complete VPC setup, RDS for databases, EFS for uploads, and an ALB for traffic management. I also implemented GitHub Actions for CI/CD to streamline our image updates.
While this setup is functioning well, the costs are starting to pile up, with each Fargate task running around $20, plus expenses for RDS and EFS. I'm contemplating a shift to running on EC2 with ECS to reduce costs. I'd like to manage multiple containers per instance but also want to ensure scalability if needed. I've thought about possibly moving to EKS, but I'm worried about the learning curve and additional costs. I'd really appreciate any advice on how to improve this architecture and keep it cost-effective. Thanks!
4 Answers
Have you thought about using a managed service like WP Engine? It can handle all your headaches around hosting and scaling, and might be cheaper in the long run!
Your journey is really impressive! It’s great to see someone who learns by doing. Fargate is awesome for managing workloads without much overhead, but I get why you’re considering EC2. Using Graviton-based t4g instances can be a game changer for performance and cost. They let you efficiently run multiple WordPress containers, plus the newer autoscaling features can optimize your resource usage. Just remember, if you go the EKS route, know it's a different beast altogether in terms of management!
Google Cloud Run is another option worth checking out if you're flexible. I’ve used it for smaller projects and it works beautifully for serverless deployments!
It sounds like you've built a pretty solid foundation! I've recently done something similar for clients migrating away from CPanel to AWS. I used Cloudpanel for easy management, which automatically mounts EFS for file storage. This way, every instance scales seamlessly, and you can keep the performance high since most traffic is handled by Varnish. It's definitely setup to avoid the overhead of CloudFront too, which can be a hassle. Besides, I have cron jobs syncing configurations — very handy! Just make sure that EFS doesn't become a bottleneck if you have multiple sites running. How are you managing the plugins? Have you considered centralizing backups too?
I'm curious about the scaling too! With multiple clients, are you able to update images in bulk without hassle?
Interesting approach! I initially hesitated to use EFS but found it really good for uploads. Considering caching solutions like Redis might be beneficial too; they help with performance quite a bit.