Hi guys!! I recently started playing around with Prefect, and I am currently trying to host my own server. So far, I've deployed the Helm Charts for the Prefect Server and Prefect Worker on a Kubernetes cluster. I would like to understand how can I submit my flows, located on my local machine, to my remote Prefect server. As from this discussion and from this tutorial, I should set the backend to

server

mode. But when I run the

prefect backend server

command, I get the following error. My prefect version is 2.10.4. Am I doing something wrong?

how can I submit my flows

This can be interpreted differently. 1. If you want the prefect server to trigger your flows , look into deployments. 2. If you want the flows you run locally to be registered on prefect server [and displayed on the UI] , Look into prefect profile.

From what I understood, a flow deployment should submit my flow to be run on the remote workers I setr up along with the Prefect Server, right? I can't clearly understand why any setup regarding profiling would change the behaviour in which my flows run.

Unsure what you mean. But a deployment just tells the infrastructure [kubernetes agent job in your case] to execute a code in specified location at given times. The server doesnt need to have access to the code . The workers on the other had should have access to the code to be executed.

Ok, but what exactly do you mean by "having access to the code"? How should I grant access to my code? From what I understood, the steps would be to: 1. Deploy a Prefect Server 2. Create a Work Pool 3. Deploy a Prefect Worker 4. Attach the worker to the pool 5. Submit my local flow to the server The fifth step is not clear to me as how I should do it. Am i getting it wrong?

You have the right steps. Now you just need to deploy your flows. Look into deployment.

Good to know that the conceptual steps are clear. But, considering that I am using a self-hosted Prefect Server, how exactly should I overwrite the default setting (that is Prefect Cloud) and use instead my own remote server? I saw that there is a

deployment.yaml

file that I can use to pre-set some parameters regarding the deployment process. Should I edit this file to point to my target server IP?

Multiple approaches to this. But you can configure your local profile to connect to your server IP https://docs.prefect.io/latest/concepts/settings/