We are using Airflow as orchestrator where it schedule workflow every hour. DataprocSubmitJobOperator is configured to schedule dataproc jobs (it uses spark). Spark sync data from source to target (runs for 50 min and then completes to avoid next schedule overlap).
Intermittent Airflow task fails due to zombie Exception. Logs show assertion failure due to pthread_mutex_lock(mu). Airflow Task exits. Underlying dataproc Job keeps running without issue.
Please suggest what can be potential issue/fix?
[2021-12-22 23:01:17,150] {dataproc.py:1890} INFO - Submitting job
[2021-12-22 23:01:17,804] {dataproc.py:1902} INFO - Job 27a2c88d-1308-4407-b965-aa490e2217fb submitted successfully.
[2021-12-22 23:01:17,805] {dataproc.py:1905} INFO - Waiting for job 27a2c88d-1308-4407-b965-aa490e2217fb to complete
E1222 23:45:58.299007027 1267 sync_posix.cc:67] assertion failed: pthread_mutex_lock(mu) == 0
[2021-12-22 23:46:00,943] {local_task_job.py:102} INFO - Task exited with return code Negsignal.SIGABRT
Config
raw_data_sync = DataprocSubmitJobOperator(
task_id="raw_data_sync",
job=RAW_DATA_GENERATION,
location='us-central1',
project_id='1f780b38bd7b0384e53292de20',
execution_timeout=timedelta(seconds=3420),
dag=dag
)
Related
We face a lot of our Airflow (MWAA) tasks receiving SIGTERM:
[2022-10-06 06:23:48,347] {{logging_mixin.py:104}} INFO - [2022-10-06 06:23:48,347] {{local_task_job.py:188}} WARNING - State of this instance has been externally set to success. Terminating instance.
[2022-10-06 06:23:48,348] {{process_utils.py:100}} INFO - Sending Signals.SIGTERM to GPID 2740
[2022-10-06 06:23:55,113] {{taskinstance.py:1265}} ERROR - Received SIGTERM. Terminating subprocesses.
[2022-10-06 06:23:55,164] {{process_utils.py:66}} INFO - Process psutil.Process(pid=2740, status='terminated', exitcode=1, started='06:23:42') (2740) terminated with exit code 1
It happens to a few of our tasks and it would not have been a big deal if the tasks were not set as a SUCCESS:
State of this instance has been externally set to success. Terminating instance
We understood that this can happen because of a lack of memory within the worker. We tried to increase the number of workers without any success. What would be our solutions to avoid having set tasks externally killed?
When tasks are getting killed, they are marked as failed. Here it seems to be the other way around. The task seem to get marked by something/someone as a success, after which the job is stopped/killed.
I am not aware of how Mwaa is deployed, but I would have a look at the action logging to see what/who is marking these tasks as success.
Hi I'm currently running airflow on a Dataproc cluster. My DAGs used to run fine but facing this issue where tasks are ending up in 'retry' state without any logs when I click on task instance -> logs on airflow UI
I see the following error in terminal where I started the airflow webserver
2022-06-24 07:30:36.544 [ERROR] Executor reports task instance
<TaskInstance: **task name** 2022-06-23 07:00:00+00:00 [queued]> finished (failed)
although the task says its queued. Was the task killed externally?
None
[2022-06-23 06:08:33,202] {models.py:1758} INFO - Marking task as UP_FOR_RETRY
2022-06-23 06:08:33.202 [INFO] Marking task as UP_FOR_RETRY
What I tried so far
restarted webserver
Started server from 3 different ports
re-ran backfill command with 3 different timestamps
deleted dag runs for my dag, created a new dag run and then re-ran backfill command
cleared the PID as mentioned here How do I restart airflow webserver? and restarted the webserver
None of these worked. This issue is persistent for the past two days, appreciate any help here.At this point I'm guessing this is to do with a shared DB but not sure how to fix this.
<<update>> So what I also found is these tasks eventually go to success or failure state. when that happens the logs are available, but still no logs for the retry attempts in $airflow_home or our remote directory
The issue was there was another celery worker listening on the same queue. since this second worker was not configured properly it was failing the task and not writing the logs to remote location.
Aiflow: 2.1.2 -
Executor: KubernetesExecutor -
Python: 3.7
I have written tasks using Airflow 2+ TaskFlow API and running the Airflow application in KubernetesExecutor mode. There are success and failure callbacks on the task but sometimes they get missed.
I've tried to specify the callbacks both via default_args on DAG and directly in the task decorator but seeing same behaviour.
#task(
on_success_callback=common.on_success_callback,
on_failure_callback=common.on_failure_callback,
)
def delta_load_pstn(files):
# doing something here
Here are the closing logs of the task
2022-04-26 11:21:38,494] Marking task as SUCCESS. dag_id=delta_load_pstn, task_id=dq_process, execution_date=20220426T112104, start_date=20220426T112131, end_date=20220426T112138
[2022-04-26 11:21:38,548] 1 downstream tasks scheduled from follow-on schedule check
[2022-04-26 11:21:42,069] State of this instance has been externally set to success. Terminating instance.
[2022-04-26 11:21:42,070] Sending Signals.SIGTERM to GPID 34
[2022-04-26 11:22:42,081] process psutil.Process(pid=34, name='airflow task runner: delta_load_pstn dq_process 2022-04-26T11:21:04.747263+00:00 500', status='sleeping', started='11:21:31') did not respond to SIGTERM. Trying SIGKILL
[2022-04-26 11:22:42,095] Process psutil.Process(pid=34, name='airflow task runner: delta_load_pstn dq_process 2022-04-26T11:21:04.747263+00:00 500', status='terminated', exitcode=<Negsignal.SIGKILL: -9>, started='11:21:31') (34) terminated with exit code Negsignal.SIGKILL
[2022-04-26 11:22:42,095] Job 500 was killed before it finished (likely due to running out of memory)
And i can see in the task instance details that the callbacks are configured.
If I implement the on_execute_callback which is called before the execution of the task, I do get the alert (in Slack). So my guess is it's definitely something with killing the pod before the callback is handled.
Our airflow installation is using CeleryExecutor.
The concurrency configs were
# The amount of parallelism as a setting to the executor. This defines
# the max number of task instances that should run simultaneously
# on this airflow installation
parallelism = 16
# The number of task instances allowed to run concurrently by the scheduler
dag_concurrency = 16
# Are DAGs paused by default at creation
dags_are_paused_at_creation = True
# When not using pools, tasks are run in the "default pool",
# whose size is guided by this config element
non_pooled_task_slot_count = 64
# The maximum number of active DAG runs per DAG
max_active_runs_per_dag = 16
[celery]
# This section only applies if you are using the CeleryExecutor in
# [core] section above
# The app name that will be used by celery
celery_app_name = airflow.executors.celery_executor
# The concurrency that will be used when starting workers with the
# "airflow worker" command. This defines the number of task instances that
# a worker will take, so size up your workers based on the resources on
# your worker box and the nature of your tasks
celeryd_concurrency = 16
We have a dag that executes daily. It has around some tasks in parallel following a pattern that senses whether the data exists in hdfs then sleep 10 mins, and finally upload to s3.
Some of the tasks has been encountering the following error:
2019-05-12 00:00:46,212 ERROR - Executor reports task instance <TaskInstance: example_dag.task1 2019-05-11 04:00:00+00:00 [queued]> finished (failed) although the task says its queued. Was the task killed externally?
2019-05-12 00:00:46,558 INFO - Marking task as UP_FOR_RETRY
2019-05-12 00:00:46,561 WARNING - section/key [smtp/smtp_user] not found in config
This kind of error occurs randomly in those tasks. When this error happens, the state of task instance is immediately set to up_for_retry, and no logs in the worker nodes. After some retries, they execute and finished eventually.
This problem sometimes gives us large ETL delay. Anyone knows how to solve this problem?
We were facing similar problems , which was resolved by
"-x, --donot_pickle" option.
For more information :- https://airflow.apache.org/cli.html#backfill
I was seeing very similar symptoms in my DagRuns. I thought it was due to the ExternalTaskSensor and concurrency issues given the queuing and killed task language that looked like this: Executor reports task instance <TaskInstance: dag1.data_table_temp_redshift_load 2019-05-20 08:00:00+00:00 [queued]> finished (failed) although the task says its queued. Was the task killed externally? But when I looked at the worker logs, I saw there was an error caused by setting a variable with Variable.set in my DAG. The issue is described here duplicate key value violates unique constraint when adding path variable in airflow dag where the scheduler polls the dagbag at regular intervals to refresh any changes dynamically. The error with every heartbeat was causing significant ETL delays.
Are you performing any logic in your wh_hdfs_to_s3 DAG (or others) that might be causing errors or delays / these symptoms?
We fixed this already. Let me answer myself question:
We have 5 airflow worker nodes. After installing flower to monitor the tasks distributed to these nodes. We found out that the failed task was always sent to a specific node. We tried to use airflow test command to run the task in other nodes and they worked. Eventually, the reason was a wrong python package in that specific node.
airflow 1.8.1
Scheduler, worker and webserver are running in separate dockers on AWS.
The system was operational, and now for some reason all tasks are staying in queued state...
No errors in scheduler logs.
In worker I see this error (not sure if its related since scheduler should move tasks from queued state):
[2018-01-23 20:46:00,428] {base_task_runner.py:95} INFO - Subtask: [2018-01-23 20:46:00,428] {models.py:1122} INFO - Dependencies not met for , dependency 'Task Instance State' FAILED: Task is in the 'success' state which is not a valid state for execution. The task must be cleared in order to be run.
I tried reboots, airflow clear and then resetdb commands but it did not help.
Any idea what else can be done to fix that problem?
Thanks