[
https://issues.apache.org/jira/browse/SPARK-37005?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
WeiNan Zhao updated SPARK-37005:
--------------------------------
Description:
when i submit a spark job , use spark-submit and set option --files file1,
then in python code, i use
{code:java}
//代码占位符
path = str(os.environ["SPARK_YARN_STAGING_DIR"])
{code}
return path is None, but this can success in java code
{code:java}
//代码占位符
spark.read().textFile(System.getenv("SPARK_YARN_STAGING_DIR") + "/README.md")
{code}
which cause this problem.
was:
hi, all
i commit a spark job , use spark-submit and set option --files file1,
then in code,i use
{code:java}
//代码占位符
path = str(os.environ["SPARK_YARN_STAGING_DIR"])
{code}
but path is None, this can success in java code
{code:java}
//代码占位符
spark.read().textFile(System.getenv("SPARK_YARN_STAGING_DIR") + "/README.md")
{code}
which cause this problem.
> pyspark os.getenv('SPARK_YARN_STAGING_DIR') can not get job path
> ----------------------------------------------------------------
>
> Key: SPARK-37005
> URL: https://issues.apache.org/jira/browse/SPARK-37005
> Project: Spark
> Issue Type: Bug
> Components: PySpark
> Affects Versions: 2.4.4
> Environment: python2.7
> spark2.4
> Reporter: WeiNan Zhao
> Priority: Major
> Labels: python, spark-core
>
> when i submit a spark job , use spark-submit and set option --files file1,
> then in python code, i use
> {code:java}
> //代码占位符
> path = str(os.environ["SPARK_YARN_STAGING_DIR"])
> {code}
> return path is None, but this can success in java code
> {code:java}
> //代码占位符
> spark.read().textFile(System.getenv("SPARK_YARN_STAGING_DIR") + "/README.md")
> {code}
> which cause this problem.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
