@Marvin is there a possibility to not store flow run parameters in Prefect? For example, I want to parse large list containing millions of integers, to flow function and I do not want to have this information stored due to memory issues

The documentation and articles available don't explicitly state a way to prevent flow run parameters from being stored within Prefect. Parameterization is a first-class feature of flows which allows for varying input to flows; however, the handling of large data being passed as parameters isn't directly addressed in the excerpts. If your concern is about avoiding large data being stored in Prefect's backend for memory optimization, you may want to consider alternative approaches such as: - Using external storage systems (like databases, object storage, or file systems) to store the large list, and passing just identifiers or references to the flow. - Investigate Prefect's approach to data management within tasks and see if a custom implementation can be made to handle large data without storing it as part of the run metadata. Unfortunately, the available documentation doesn't provide a direct answer to this query, so you might want to consider reaching out to the Prefect community or consulting the official documentation for more detailed guidance on handling large data sets. If Prefect 2.x has different mechanisms or approaches, further research or questions directed to the community may yield better insights. For further details, you may want to review the provided links from the documentation: - Flows in Prefect - Data is Mail, Not Water - Blog Post

Yeah, for things like this we recommend doing something like storing that data externally and passing the path to that storage as your parameter, for the data to then be loaded and processed in the context it's needed