[ 
https://issues.apache.org/jira/browse/AIRFLOW-1319?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Chris Sng updated AIRFLOW-1319:
-------------------------------
    Description: 
In the community-contributed Spark submit hook and operator, it support 
{{spark-submit}}'s {{--files}} command line option. The {{--files}} option is 
used to submit file to each executor to be used. A good example of such files 
are serialized objects.

However, in both docstrings 
(https://github.com/apache/incubator-airflow/blob/master/airflow/contrib/operators/spark_submit_operator.py#L37
 and 
https://github.com/apache/incubator-airflow/blob/master/airflow/contrib/hooks/spark_submit_hook.py#L36),
 it provided {{hive-site.xml}} as an example. This may mislead less-informed 
developers into assuming that Hive configuration files can be submitted to the 
cluster in this manner. According to Apache Hive's documentation, hive 
configuration files are located in the directory located in the `HIVE_CONF_DIR` 
environment variable.

I propose excluding this example from the docstrings.

  was:
In the community-contributed Spark submit hook and operator, it support 
`spark-submit`'s `--files` command line option. The `--files` option is used to 
submit file to each executor to be used. A good example of such files are 
serialized objects.

However, in both docstrings 
(https://github.com/apache/incubator-airflow/blob/master/airflow/contrib/operators/spark_submit_operator.py#L37
 and 
https://github.com/apache/incubator-airflow/blob/master/airflow/contrib/hooks/spark_submit_hook.py#L36),
 it provided `hive-site.xml` as an example. This may mislead less-informed 
developers into assuming that Hive configuration files can be submitted to the 
cluster in this manner. According to Apache Hive's documentation, hive 
configuration files are located in the directory located in the `HIVE_CONF_DIR` 
environment variable.

I propose excluding this example from the docstrings.


> Fix misleading SparkSubmitOperator and SparkSubmitHook docstring
> ----------------------------------------------------------------
>
>                 Key: AIRFLOW-1319
>                 URL: https://issues.apache.org/jira/browse/AIRFLOW-1319
>             Project: Apache Airflow
>          Issue Type: Improvement
>          Components: contrib
>            Reporter: Chris Sng
>            Assignee: Chris Sng
>            Priority: Trivial
>              Labels: documentation
>   Original Estimate: 5m
>  Remaining Estimate: 5m
>
> In the community-contributed Spark submit hook and operator, it support 
> {{spark-submit}}'s {{--files}} command line option. The {{--files}} option is 
> used to submit file to each executor to be used. A good example of such files 
> are serialized objects.
> However, in both docstrings 
> (https://github.com/apache/incubator-airflow/blob/master/airflow/contrib/operators/spark_submit_operator.py#L37
>  and 
> https://github.com/apache/incubator-airflow/blob/master/airflow/contrib/hooks/spark_submit_hook.py#L36),
>  it provided {{hive-site.xml}} as an example. This may mislead less-informed 
> developers into assuming that Hive configuration files can be submitted to 
> the cluster in this manner. According to Apache Hive's documentation, hive 
> configuration files are located in the directory located in the 
> `HIVE_CONF_DIR` environment variable.
> I propose excluding this example from the docstrings.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to