I think you’d have a little more luck asking on something like stackoverflow,
as it seems most devs are preoccupied by the outstanding issues and preparing a
release candidate.
I’m not sure that the default ts_function is your issue, but I think that if
you’re trying to use a custom one, you’ll need to implement it as a function:
https://github.com/apache/incubator-airflow/blob/3fde1043fe56fb0772d0c0f2a1b32105b18f2583/airflow/contrib/sensors/gcs_sensor.py#L102
So your
ts_func = '{{ execution_date}}'
could be written as
ts_func=lambda: context['execution_date']
-Daniel
On 10/31/17, 1:01 AM, "Sam Elamin" <[email protected]> wrote:
Hi Everyone
does anyone have any problems creating a
GoogleCloudStorageObjectUpdatedSensor
<https://github.com/apache/incubator-airflow/blob/master/airflow/contrib/sensors/gcs_sensor.py>
?
im on airflow 1.8 and i am getting an error of
*File
"/usr/local/lib/python2.7/dist-packages/airflow/contrib/sensors/gcs_sensor.py",
line 73, in ts_function*
*[2017-10-30 15:27:10,609] {base_task_runner.py:95} INFO - Subtask: return
context['execution_date'] + context['dag'].schedule_interval*
*[2017-10-30 15:27:10,609] {base_task_runner.py:95} INFO - Subtask:
TypeError: unsupported operand type(s) for +: 'datetime.datetime' and 'str'*
*[2017-10-30 15:27:15,906] {jobs.py:2097} INFO - Task exited with return
code 1*
i am not overwriting the ts_function so it seems the default ts_function
isn't working
When I try to overwrite the dag with
ts_fun = '{{ ds }}' or ts_func = '{{ execution_date}}'
I get
*''str' object is not callable'*
Which I assume is because its reading the date as a string instead of a
datetime object
Any help would be greatly apprecated?
Regards
Sam
