Hello,
Do you plan to deploy our apps in managed containers instead of Kubernetes?
Like ECS, Cloud Run or Scaleway Serverless Containers.
Why? For two reasons:
Sometimes, Kubernetes stay overkill for simple backend+fronted.
Kubernetes can be a limit that some organisations are not ready to cross. Some startups keep Heroku even if the UI is terrible for multi-environment (imagine 7 apps in 6 envs). Even if Heroku is more expensive, they say this is the price for do not hire DevOps. If you speak with them about Kubernetes, you already lost … despit Qovery product perfectly matches their philosophy.
With this feature, I believe Qovery will become the standard for deploying any code anywhere.
Thanks for your message. Actually Qovery provides an option where the Kubernetes cluster is fully managed by Qovery and you don’t require any Kubernetes knowledge. Did you explore this option? If yes, what are your concerns?
The logic is even if Qovery manage the cluster, you still work in a Kubernetes world:
Instances, nodes, auto scale nodes, CPU and Mem request and limit, etc
In Heroku and managed containers, that concepts do not exists
I got your point, and I’m glad to have the opportunity to clarify our positioning
Our vision is to turn every single developer into their own DevOps engineers. It does not mean that we want to fully abstract the infrastructure pieces, but instead provide an outstanding Developer Experience on top of it without assuming that the developer understands anything about infrastructure. The ultimate goal for us is to make sure that it’s so obvious for developers to deploy and run their apps with Qovery that they don’t have to think about the infrastructure. But at the same time, we want to give you the full power of the cloud/infrastructure you are using. It’s a big challenge because we always need to arbitrate between the simplicity and the flexibility. That’s why we have designed some patterns like Advanced Settings and BYOK.
Those configuration elements are disappearing - we just released GCP GKE Autopilot, and for instance, you no longer need to configure your cluster nodes. It’s coming with our AWS EKS offer as well for Q2 2024
For your application, you can still do like on Heroku and over-allocate resources. Still, it’s a waste of money for you, but it works and helps you get started faster and not bother about what the best allocation is for you. But that’s something that we will ultimately help at some point. Responding to the question: what’s the best allocation for your app.
I hope it helps to better understand our philosophy
