PrefectHQ / prefect

Prefect is a workflow orchestration framework for building resilient data pipelines in Python.
https://prefect.io
Apache License 2.0
16.01k stars 1.57k forks source link

Customize Run Environment variables #6470

Closed BitTheByte closed 2 years ago

BitTheByte commented 2 years ago

First check

Prefect Version

2.x

Describe the proposed behavior

Add the ability to customize run environment variables from UI even if the deployment has an infra block attached to it

Describe the current behavior

Not Implemented

Example Use

No response

Additional context

No response

anna-geller commented 2 years ago

@BitTheByte, what you are asking for is currently not possible by using environment variables (i.e., you cannot change the environment variable for a specific run), however, you can solve the problem by putting the same information into a parameter and you can totally start a custom parametrized run from the UI. Here is what this could look like: image

In this example, user could be the same information that you currently have stored in your environment variable.

Given that this can already be solved, I'll close the issue and if this doesn't solve the problem or if I misunderstood the problem, please reopen the issue and describe it in more detail or create a separate more detailed issue. Thanks a lot!

BitTheByte commented 2 years ago

@anna-geller Thanks for your detailed reply, however I wanted to customize environment variables as I wanted to configure task runner options using it. Also some libraries may require their configuration to be passed using environment variables I don't wish to have a dozen of deployments for the same workflow just bc I wanted to change some configurations. I'll keep the issue closed. Please feel free to reopen it if the solution can be implemented in the future.