Hi Ram, You can have a single DAG scheduled to 10am, which starts A and then use a TimeSensor set to 11 am that B depends on and then have C depend on A and B.
Something like: a = BashOperator(‘a’, …) delay_b = TimeSensor(‘delay_b’, target_time=time(11, 0, 0), …) b = BashOperator(‘b’, …) b.set_upstream(delay_b) c = BashOperator(‘c’, …) c.set_upstream(a) c.set_upstream(b) / Carl Johan On 23 July 2018 at 02:18:00, srinivas.ramabhad...@gmail.com (srinivas.ramabhad...@gmail.com) wrote: Hi - I have recently started using Airflow version 1.9.0 and am having some difficulty setting up a very simple DAG. I have three tasks A, B and C. I'd like A to run every day at 10am and B at 11am. C depends on BOTH A and B running successfully. Initially, I decided to create one DAG, add all three tasks to it and set C as downstream to A and B. I then set the schedule_interval of the DAG to @daily. But this meant I couldn't run A and B at 10am and 11am respectively since the they are PythonOperators and tasks dont support schedule_interval (or, at least, it's deprecated syntax and gets ignored). I scratched that idea and then created A and B as DAGs, specified the schedule interval as per the cron syntax: '00 10 * * *' for A and '00 11 * * *' for B. But now when I set C as a downstream of A and B, it complains that C can't belong to two different dags. How do I accomplish such a simple dependency structure? Ram.
