Skip to content

Status of testing Providers that were prepared on February 08, 2023 #29424

Closed

Description

@eladkal

Body

I have a kind request for all the contributors to the latest provider packages release.
Could you please help us to test the RC versions of the providers?

Let us know in the comment, whether the issue is addressed.

Those are providers that require testing as there were some substantial changes introduced:

Provider amazon: 7.2.0rc1

Provider apache.beam: 4.2.0rc1

Provider apache.hive: 5.1.2rc1

Provider arangodb: 2.1.1rc1

Provider cncf.kubernetes: 5.2.0rc1

Provider dbt.cloud: 3.0.0rc1

Provider elasticsearch: 4.4.0rc1

Provider ftp: 3.3.1rc1

Provider google: 8.9.0rc1

Provider microsoft.azure: 5.2.0rc1

Provider mysql: 4.0.1rc1

Provider presto: 4.2.2rc1

Provider sftp: 4.2.2rc1

Provider snowflake: 4.0.3rc1

Provider tableau: 4.1.0rc1

Provider trino: 4.3.2rc1

The guidelines on how to test providers can be found in

Verify providers by contributors

Contributors are also encouraged to cast vote in the mailing list thread.

All users involved in the PRs:
@BasPH @VladaZakharova @bkossakowska @pankajkoti @phanikumv @MaksYermak @RachitSharma2001 @vandonr-amz @IAL32 @raphaelauv @Taragolis @dwreeves @potiuk @snjypl @dstandish @vchiapaikeo @eladkal @shahar1
@jedcunningham @csm10495 @josh-fell @joarobles @johannaojeling @o-nikolas @wonhyeongseo @seunggabi @MrGeorgeOwl @hussein-awala @maggesssss

Committer

  • I acknowledge that I am a maintainer/committer of the Apache Airflow project.

Activity

  1. added
    kind:metaHigh-level information important to the community
    on Feb 8, 2023
  2. pankajkoti commented on Feb 9, 2023

    @pankajkoti
    Member

    Thanks @eladkal . I tested my PR #29349 which is part of Provider google: 8.9.0rc1 and it works fine.

  3. seunggabi commented on Feb 9, 2023

    @seunggabi
    Contributor

    Thanks @eladkal .
    I tested a minute ago.

    #29360: @seunggabi
    https://pypi.org/project/apache-airflow-providers-tableau/4.1.0rc1/

    It was works.

    from airflow.providers.tableau.operators.tableau import TableauOperator
    
    ...
    
    task_refresh_workbook_blocking = TableauOperator(
        resource="workbooks",
        method="refresh",
        match_with="{{ params.match_with }}",
        find="{{ params.find }}",
        blocking_refresh=True,
        task_id="refresh_tableau_workbook_blocking",
        dag=dag,
    )
    

    스크린샷 2023-02-10 오전 12 11 31

  4. johannaojeling commented on Feb 9, 2023

    @johannaojeling
    Contributor

    #28764 is tested with the Beam RC and works as expected

  5. csm10495 commented on Feb 9, 2023

    @csm10495
    Contributor

    #28808 still seems legit

  6. vchiapaikeo commented on Feb 9, 2023

    @vchiapaikeo
    Contributor

    #28959 looks good

    write_on_empty=False (default)
    image

    write_on_empty=True
    image

  7. o-nikolas commented on Feb 9, 2023

    @o-nikolas
    Contributor

    I added code to kill the process while running to test the emr on_kill changes, and the cleanup code ran successfully (and I confirmed the resource was deleted in my AWS account):

    INFO  [airflow.task.operators] Job flow with id j-1ADYKTIV4THO5 created
    INFO  [airflow.task.operators] SLEEPING NOW, KILL THE PROCESS TO TEST ON_KILL
    INFO  [airflow.task.operators] TRYING TO KILL MYSELF
    ERROR [airflow.task] Received SIGTERM. Terminating subprocesses.
    INFO  [airflow.task.operators] Terminating job flow j-1ADYKTIV4THO5
    

    While I was there, I tested the wait_for_completion changes on the same operator. It worked successfully:

    INFO  [airflow.task] Executing <Task(EmrCreateJobFlowOperator): create_job_flow> on 2021-01-01 00:00:00+00:00
    INFO  [airflow.task] Exporting env vars: AIRFLOW_CTX_DAG_OWNER='airflow' AIRFLOW_CTX_DAG_ID='example_emr' AIRFLOW_CTX_TASK_ID='create_job_flow' AIRFLOW_CTX_EXECUTION_DATE='2021-01-01T00:00:00+00:00' AIRFLOW_CTX_TRY_NUMBER='1' AIRFLOW_CTX_DAG_RUN_ID='backfill__2021-01-01T00:00:00+00:00'
    INFO  [airflow.task.operators] Creating job flow using aws_conn_id: aws_default, emr_conn_id: emr_default
    INFO  [airflow.hooks.base] Using connection ID 'emr_default' for task execution.
    INFO  [airflow.hooks.base] Using connection ID 'aws_default' for task execution.
    INFO  [botocore.credentials] Found credentials in environment variables.
    INFO  [airflow.task.operators] Job flow with id j-2VDGDHAR3ASCT created
    INFO  [airflow.providers.amazon.aws.utils.waiter] Waiting for job flow to be finished.
    INFO  [airflow.providers.amazon.aws.utils.waiter] Waiting for job flow to be finished.
    INFO  [airflow.providers.amazon.aws.utils.waiter] Waiting for job flow to be finished.
    INFO  [airflow.providers.amazon.aws.utils.waiter] Waiting for job flow to be finished.
    INFO  [airflow.providers.amazon.aws.utils.waiter] Waiting for job flow to be finished.
    

    I marked both changes as tested above in the task list.

  8. josh-fell commented on Feb 10, 2023

    @josh-fell
    Contributor

    Thanks for organizing @eladkal!

    All of the dbt Cloud changes look good as well as the Python API docs change in #29041.

  9. eladkal commented on Feb 11, 2023

    @eladkal
    ContributorAuthor

    Thank you everyone.
    Providers are released
    I invite everyone to help improve providers for the next release, a list of open issues can be found here.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    kind:metaHigh-level information important to the communitytesting statusStatus of testing releases

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions