You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Keith Sun (JIRA)" <ji...@apache.org> on 2017/08/23 10:55:00 UTC

[jira] [Created] (SPARK-21819) UserGroupInformation initialization in SparkHadoopUtilwill overwrite user config

Keith Sun created SPARK-21819:
---------------------------------

             Summary:  UserGroupInformation initialization in SparkHadoopUtilwill overwrite user config
                 Key: SPARK-21819
                 URL: https://issues.apache.org/jira/browse/SPARK-21819
             Project: Spark
          Issue Type: Bug
          Components: Java API, Spark Core
    Affects Versions: 2.1.1, 2.1.0
         Environment: Ubuntu14.04
Spark2.10/2.11 (I checked the github of 2.20 , it exist there as well)

            Reporter: Keith Sun


When  submit job in Java or Scala code to ,the initialization of SparkHadoopUtil will trigger the configuration overwritten in UGI which may not be expected if the UGI has already been initialized by customized xmls which are not on the classpath (like the cfg4j , which could set conf from github code, a database etc)

{code:java}
//it will overwrite the UGI conf which has already been initialized
class SparkHadoopUtil extends Logging {
  private val sparkConf = new SparkConf(false).loadFromSystemProperties(true)
  val conf: Configuration = newConfiguration(sparkConf)
  UserGroupInformation.setConfiguration(conf)
{code}





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

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org