You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@airflow.apache.org by "Daniel Standish (JIRA)" <ji...@apache.org> on 2019/03/10 01:13:00 UTC
[jira] [Created] (AIRFLOW-4056) Dag file processing does not
respect dag_dir_list_interval
Daniel Standish created AIRFLOW-4056:
----------------------------------------
Summary: Dag file processing does not respect dag_dir_list_interval
Key: AIRFLOW-4056
URL: https://issues.apache.org/jira/browse/AIRFLOW-4056
Project: Apache Airflow
Issue Type: Bug
Components: scheduler
Affects Versions: 1.10.2
Environment: I have confirmed this issue on mac and centos environments, using mysql backend.
Reporter: Daniel Standish
The conf parameter dag_dir_list_interval seems to have no effect on dag directory scanning.
It seems to happen every 2 seconds, no matter what. The default is supposed to be 5 minutes.
As a result I see a scheduler output like this:
{code:java}
[2019-03-09 17:06:24,579] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:26,587] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:28,590] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:30,597] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:32,603] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:34,611] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:35,195] {sqlalchemy.py:79} WARNING - DB connection invalidated. Reconnecting...
[2019-03-09 17:06:36,615] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:38,623] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:40,631] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:42,637] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:44,644] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:46,205] {sqlalchemy.py:79} WARNING - DB connection invalidated. Reconnecting...
[2019-03-09 17:06:46,651] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:48,658] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:50,666] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:52,670] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:54,680] {jobs.py:1559} INFO - Harvesting DAG parsing results
[2019-03-09 17:06:56,687] {jobs.py:1559} INFO - Harvesting DAG parsing results{code}
And no more is there the periodic printing of dag stats, like there was in 1.10.1.
I can confirm that this is happening by adding this to something in dag folder:
{code:java}
with open(Path('~/temp/test.log').expanduser(), 'at') as f:
f.write(f"{datetime.now()}: i am imported\n")
{code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)