Planning a migration between Airflow environments
As your Airflow instances grow, it is not uncommon to wish to migrate workloads to new environments. Sometimes, this is due to a change in service providers; other times, it is due to a need to share your workflows so that teams/workloads that once ran in the same deployment are isolated from each other.
All of our previous advice still holds true here, but because you’re going from Airflow to Airflow, it’s a generally much more straightforward activity and may not require as much testing to complete.
After identifying the DAGs you wish to migrate, you need to identify which objects you need to migrate to support their execution and whether you need to migrate previous execution history.
Connections and variables
Identify any connection or variable objects that you need for your DAG; if you’re using a secrets backend, you will need to consult with that service for how to migrate variables to a new environment...
