from airflow import DAG
from airflow.operators.bash_operator import BashOperator
from datetime import datetime, timedelta
from airflow.contrib.operators.spark_submit_operator import SparkSubmitOperator
from datetime import timedelta,datetime
default_args = {
'owner': 'Terry',
'depends_on_past': False,
'start_date': "2021-08-02",
'email': ['tian23.li@tcl.com'],
'email_on_failure': True,
'email_on_retry': False,
'retries': 1,
'retry_delay': timedelta(minutes=30),
}
dag = DAG(
dag_id='remove_history_files',
description='This dag is use to remove history files',
default_args=default_args,
schedule_interval='0 3 * * *',
)
# remove_history_files = BashOperator(
# task_id='remove_history_files',
# bash_command="sh /data/gangyanyuan/rm_history_files.sh ",
# dag=dag,
# )
remove_history_files_2024 = BashOperator(
task_id='remove_history_files_2024',
bash_command="sh /data/gangyanyuan/rm_history_files_2024.sh ",
dag=dag,
)
# remove_history_files_gary = BashOperator(
# task_id='remove_history_files_gary',
# bash_command="cd /data/gangyanyuan/gary && bash remove_unused_files.sh ",
# dag=dag,
# )
remove_history_files_2024