I've created a Kubernetes lab setup that I'm pretty proud of, featuring an architecture of 3 masters and 2 workers with HA configured using Ansible. Currently, my infrastructure consists of 6 VMs powered by KVM/QEMU, and I've integrated several tools like Falco, Grafana, Prometheus, and Trivy. However, I'm running low on disk space. Since my PC only has one slot, I'm considering rebuilding the environment from scratch on Proxmox, utilizing Terraform for VM creation and Ansible for configuration. I'd love to get your thoughts on this plan! Here are my specific queries:
1. How much time do you think it would take to recreate this entire setup with Terraform and Ansible?
2. What are your suggestions for memory and disk space allocations for each VM (both masters and workers) to maintain good performance for this lab?
3. Are there any tips, best practices, or "gotchas" I should bear in mind while transitioning to Proxmox and Terraform for my Kubernetes lab?
1 Answer
Have you considered consolidating your VMs? You might be able to combine your control nodes with your working nodes, possibly running everything on just 3 or 4 VMs. Since you're using a single machine, having all those extra VMs might not be necessary. Also, sharing the operating system disks and using the same base image across your VMs can save you some space. QCOW2 supports backing base files which is super helpful. If you're mainly testing Kubernetes workloads, tools like kind might make your setup easier and resource-efficient.
You're right! If the goal is just to run Kubernetes workloads and test basic deployments, then a simpler setup could work perfectly. But since this is your university project focusing on simulating a real-world cluster with all the DevSecOps tooling, I understand why you chose a multi-node setup to create realistic scenarios. Just keep an eye on your disk usage; template cloning with QCOW2 base images would definitely be something worth implementing!