You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@airflow.apache.org by GitBox <gi...@apache.org> on 2021/04/14 21:16:01 UTC
[GitHub] [airflow] ctimbreza opened a new issue #15374: Clearing a subdag task leaves parent dag in the failed state
ctimbreza opened a new issue #15374:
URL: https://github.com/apache/airflow/issues/15374
**Apache Airflow version**
2.0.1
**Kubernetes version**
Server Version: version.Info{Major:"1", Minor:"18+", GitVersion:"v1.18.9-eks-d1db3c", GitCommit:"d1db3c46e55f95d6a7d3e5578689371318f95ff9", GitTreeState:"clean", BuildDate:"2020-10-20T22:18:07Z", GoVersion:"go1.13.15", Compiler:"gc", Platform:"linux/amd64"}
**What happened**:
Clearing a failed subdag task does not automatically set the state of the parent dag to 'running' so that the downstream parent tasks can execute.
To reproduce:
- Clear a failed subdag task selecting option for Downstream+Recursive
- See that all down stream tasks in the subdag as well as the parent dag have been cleared
- See that the parent dag is left in 'failed' state.
Work around:
Manually set the state of the parent dag to running after clearing the subdag task
**What you expected to happen**:
With airflow version 1.10.4 the parent dag was automatically set to 'running' for this same scenario
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] jedcunningham commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
jedcunningham commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-819870840
@ctimbreza, I wasn't able to reproduce this in master naively. Is there anything in the webserver log that might be helpful? Do you have a simple DAG that reproduces it?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] ephraimbuddy commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
ephraimbuddy commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-820160136
@ctimbreza looks like this issue is resolved in https://github.com/apache/airflow/pull/14776, can you verify?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] zarrarrana commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
zarrarrana commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-827966516
@ctimbreza @ephraimbuddy
#14776 was very close in fixing this, it is missing parent dag_id to be added to dag_ids if include_parentdag is true, to change the DagRun state.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] ctimbreza commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
ctimbreza commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-820649212
@ephraimbuddy Thanks for pointing me to #14776. Looks like this is targeted for release 2.0.2 which is not out yet. I spent a little time today to see if I could run airflow locally to verify and was looking at the breeze development environment.
So far it has been pretty straightforward to setup, however after launching the environment (breeze start-airflow) and accessing the UI (http://127.0.0.1:28080/home) I cannot seem to interact with the UI. My dag is shown but none of the UI links are working.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] ctimbreza commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
ctimbreza commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-827689509
Hi folks. Looks like [#14776](https://github.com/apache/airflow/pull/14776) did not resolve this issue. I am still able to reproduce it on Airflow 2.0.2
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] ctimbreza commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
ctimbreza commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-820696636
I tried running ```cd airflow/www; yarn build prod/dev``` but received an error that I forgot to capture.
I ended up deleting the .build dir and starting over. This time when running breeze I noticed a message saying to run './airflow/www/compile_assets.sh'. I ran this and then the UI was up and responding as expected.
Hitting a new issue now where my dag doesn't seem to be processing the tasks and I am seeing this message about the scheduler not running. I started with 'breeze start-airflow' and the scheduler looks like it is running from the console window.
Thanks everyone for the quick responses on this issue. I'll need to put this aside for now as I work on higher priority items on my side. Feel free to close the issue if you think it has already been addressed.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] ephraimbuddy commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
ephraimbuddy commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-820652825
> @ephraimbuddy Thanks for pointing me to #14776. Looks like this is targeted for release 2.0.2 which is not out yet. I spent a little time today to see if I could run airflow locally to verify and was looking at the breeze development environment.
>
> So far it has been pretty straightforward to setup, however after launching the environment (breeze start-airflow) and accessing the UI (http://127.0.0.1:28080/home) I cannot seem to interact with the UI. My dag is shown but none of the UI links are working.
Probably there's no user, you can create a user in breeze environment, you just have to run the create user command:
`airflow users create -f firstname -l lastname -u admin -p admin -r Admin `
You can drop into breeze environment with :
`./breeze`
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] ctimbreza commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
ctimbreza commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-819907856
Here is the dag that I used to reproduce the issue. Subdag_1 task_2 is configured to randomly fail to help setup the scenario for this issue.
```
from airflow.models.dag import DAG
from airflow.operators.bash import BashOperator
from airflow.operators.dummy import DummyOperator
from airflow.operators.python_operator import PythonOperator
from airflow.operators.subdag_operator import SubDagOperator
from airflow.utils.dates import days_ago
import random
access_control = {
"dst-view": ["can_read"],
"dst-operator": ["can_read", "can_edit"],
"dst-admin": ["can_read", "can_edit"]
}
def random_fail(**context):
if random.random() > 0.5:
raise Exception("Test failure")
print("Success")
def __build_subdag1(parent_dag, child_dag_name):
with DAG(dag_id='%s.%s' % (parent_dag.dag_id, child_dag_name),
default_args=parent_dag.default_args,
schedule_interval=parent_dag.schedule_interval,
start_date=parent_dag.start_date,
params=parent_dag.params,
access_control=parent_dag.access_control,
is_paused_upon_creation=False) as subdag:
task_1 = DummyOperator(task_id="task_1")
task_2 = PythonOperator(task_id="task_2", python_callable=random_fail, provide_context=True)
task_1 >> task_2
return subdag
def __build_subdag2(parent_dag, child_dag_name):
with DAG(dag_id='%s.%s' % (parent_dag.dag_id, child_dag_name),
default_args=parent_dag.default_args,
schedule_interval=parent_dag.schedule_interval,
start_date=parent_dag.start_date,
params=parent_dag.params,
access_control=parent_dag.access_control,
is_paused_upon_creation=False) as subdag:
task_1 = DummyOperator(task_id="task_1")
task_2 = BashOperator(task_id="task_2", bash_command='sleep 15')
task_1 >> task_2
return subdag
with DAG(dag_id="chris-test", start_date=days_ago(2), tags=["dev"], access_control=access_control) as dag:
start = DummyOperator(task_id="start")
subdag_1 = SubDagOperator(
subdag=__build_subdag1(parent_dag=dag,
child_dag_name="subdag_1"
),
task_id="subdag_1",
dag=dag
)
subdag_2 = SubDagOperator(
subdag=__build_subdag2(parent_dag=dag,
child_dag_name="subdag_2"
),
task_id="subdag_2",
dag=dag
)
end = DummyOperator(task_id='end')
start >> subdag_1 >> subdag_2 >> end
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] xinbinhuang edited a comment on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
xinbinhuang edited a comment on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-819940226
Hi @ctimbreza, welcome to Airflow! Before solving your specific issue with SubDagOperator, I wonder if you have considered using [`TaskGroup`](https://airflow.apache.org/docs/apache-airflow/stable/concepts.html#taskgroup) for your use case instead?
It was introduced as an alternative to SubDagOperator. It's simple to use and without the shortcomings from SubDag
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] boring-cyborg[bot] commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
boring-cyborg[bot] commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-819837937
Thanks for opening your first issue here! Be sure to follow the issue template!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] ctimbreza commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
ctimbreza commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-820648096
@xinbinhuang Thanks for your suggestion on TaskGroup. We are look at TaskGroup for new dags going forward. However we do have a lot of existing dags currently running on airflow1 that we would like to move over to airflow2 as is.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] xinbinhuang commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
xinbinhuang commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-819940226
Hi @ctimbreza, welcome to Airflow! Before solving your specific issue with SubDagOperator, I wonder if you have considered using [`TaskGroup`](https://airflow.apache.org/docs/apache-airflow/stable/concepts.html#taskgroup) for your use case instead?
Though nothing official yet, the community may deprecate `SubDagOperator` in the future.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] kaxil closed issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
kaxil closed issue #15374:
URL: https://github.com/apache/airflow/issues/15374
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] xinbinhuang commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
xinbinhuang commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-820661652
> @ephraimbuddy Thanks for pointing me to #14776. Looks like this is targeted for release 2.0.2 which is not out yet. I spent a little time today to see if I could run airflow locally to verify and was looking at the breeze development environment.
>
> So far it has been pretty straightforward to setup, however after launching the environment (breeze start-airflow) and accessing the UI (http://127.0.0.1:28080/home) I cannot seem to interact with the UI. My dag is shown but none of the UI links are working.
It's likely that the frontend assets haven't been build (gone into this issue recently). Can you try, inside breeze, `cd airflow/www; yarn build prod/dev`? And then start the webserver again
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [airflow] xinbinhuang commented on issue #15374: Clearing a subdag task leaves parent dag in the failed state
Posted by GitBox <gi...@apache.org>.
xinbinhuang commented on issue #15374:
URL: https://github.com/apache/airflow/issues/15374#issuecomment-820719653
> I tried running `cd airflow/www; yarn build prod/dev` but received an error that I forgot to capture.
I mean either `cd airflow/www; yarn build prod` or `cd airflow/www; yarn build dev`. sorry for the confusion. But `comiple_assets.sh` will work too.
> Hitting a new issue now where my dag doesn't seem to be processing the tasks and I am seeing this message about the scheduler not running. I started with 'breeze start-airflow' and the scheduler looks like it is running from the console window.
You need to start both the `scheduler` and `webserver`
```
# start scheduler in background; `&` means run the command in background
airflow scheduler &
# start webserver
airflow webserver
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org