Hey everyone,

Normally, in all my previous projects, I have used a combination of cloud run (or similar) and managed compute services (ec2, compute engine, etc.) to deploy my services and this works great.

With my new project, I have deployed a prototype on a managed compute service and this works fine. However, to scale my solution, I will need to programmatically scale up persistent services with permanent storage volumes. This is not a requirement I’ve had before in previous projects and I’m coming to you all with advice on how to proceed.

This seems to be exactly what k8s is for. If I take the time to learn how to use Kubernetes, I’m sure I could use it to solve my problem.

However, I also feel confident that I can build my own solution to auto scale up these persistent containers.

After some, admittedly brief, research on how k8s works, I’m beginning to realize that learning Kubernetes will be a task in itself and will take away much of my valuable free time I could be spending on building out the product further. However, I’m worried that if I go my own path, it will inevitably lead to scaling issues in the future.

What do you guys think, spend the time to do it properly, or build a quick and dirty solution for now and worry later?