To perform an operation similar to an Airflow backfill in Prefect, you would need to create historical flow runs or rerun past executions. Unfortunately, the search did not return a straightforward guide on performing backfills. However, you can manually trigger flow runs for past dates or handle reruns of past executions using Prefect's client library.
You can schedule your flows to run for past dates by specifying the desired past date when creating a new flow run. To rerun past executions, you may consider using the Prefect client to reschedule late flow runs or create new flow runs based on past deployments.
Here are some relevant details I found in the Prefect documentation:
- Using the
Prefect Client, you can create functions to reschedule late flow runs (
reschedule_late_flow_runs
).
- You can also filter and delete flow runs, then create new ones from past deployments.
I suggest exploring the Prefect Client API to perform these tasks, as it appears to be the most suitable approach for backfill-like operations in Prefect. Here are some related documentation links that might be helpful:
-
Using the Prefect Client
-
Dbt Cloud Integration Examples
-
Prefect Databricks Jobs Models
While these sources do not offer a direct solution, they provide a starting point for understanding how to manage flow runs manually, which may be adapted for backfilling purposes.