[ 
https://issues.apache.org/jira/browse/AIRFLOW-43?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15269455#comment-15269455
 ] 

Jeremiah Lowin commented on AIRFLOW-43:
---------------------------------------

I believe the issue is that your {{schedule_interval}} is {{None}}, so the 
scheduler is ignoring your DAG. [~sanand] uses trigger_dag a lot, perhaps he 
can confirm. 

This speaks to some confusion between the Scheduler and externally-triggered 
DagRuns that [~bolke] and I are actively working on. I think setting your 
schedule_interval might help here but it's definitely an odd case for us to 
keep in mind (you don't want the DAG to be scheduled regularly... but it has to 
be executed by the Scheduler!). 

> Tasks never run in externally triggered DAGRun
> ----------------------------------------------
>
>                 Key: AIRFLOW-43
>                 URL: https://issues.apache.org/jira/browse/AIRFLOW-43
>             Project: Apache Airflow
>          Issue Type: Bug
>    Affects Versions: Airflow 1.7.1
>         Environment: OSX, Python 2.7.5 ... Airflow is 1.7.0 (not available in 
> dropdown above)
>            Reporter: Chris Johnson
>            Priority: Critical
>
> I have a DAG which should only be run as needed (no schedule interval):
> dag = DAG(
>     dag_id='major_integ_cycle',
>     default_args=args,
>     schedule_interval=None,
> )
> I attempted to start this using `airflow trigger_dag NAME` but the tasks 
> never run.  The scheduler has noted the fact that itthe DAG was triggered: 
> INFO - Created <DagRun major_integ_cycle @ 2016-05-03 08:36:16.032474: 
> manual__2016-05-03T08:36:16.032474, externally triggered: True>
> The tasks are currently very vanilla test cases, all look like this:
> task_mdc = op.BashOperator(
>     task_id='master_data_cycle',
>     bash_command='sleep 1',
>     dag=dag,
> )
> Have I misunderstood?  Or is not running a bug in this case?



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to