airflow_supervisor.SupervisorAirflowConfiguration

pydantic model airflow_supervisor.SupervisorAirflowConfiguration

Bases: SupervisorConvenienceConfiguration

Settings that MUST be set when running in airflow

field check_interval: timedelta = datetime.timedelta(seconds=5)

Interval between supervisor program status checks

field check_timeout: timedelta = datetime.timedelta(seconds=28800)

Timeout to wait for supervisor program status checks

field runtime: timedelta | None = None

Max runtime of Supervisor job

field endtime: time | None = None

End time of Supervisor job

field maxretrigger: int | None = None

Max number of retriggers of Supervisor job (e.g. max number of checks separated by check_interval)

field reference_date: Literal['start_date', 'logical_date', 'data_interval_end'] = 'data_interval_end'

Reference date for the job. NOTE: Airflow schedules after end of date interval, so data_interval_end is the default

field pool: str | Pool | None = None

Other Airflow Configuration

Airflow pool to use for the job. If not set, the job will use the default pool, or the pool from a balancer host.

field stop_on_exit: bool | None = True

Stop supervisor on dag completion

field cleanup: bool | None = True

Cleanup supervisor folder on dag completion. Note: stop_on_exit must be True

field restart_on_initial: bool | None = False

Restart the job when the DAG is run directly via airflow (NOT retriggered). This is useful for jobs that do not shutdown

field restart_on_retrigger: bool | None = False

Restart the job when the DAG is retriggered. This is useful for jobs that do not shutdown