[ https://issues.apache.org/jira/browse/SPARK-2459?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14061981#comment-14061981 ]
Nan Zhu commented on SPARK-2459: -------------------------------- yeah, spark-submit may solve the problem... then we need to modify the start-thriftserver.sh (https://github.com/apache/spark/commit/8032fe2fae3ac40a02c6018c52e76584a14b3438#diff-acab5881e22c8120bd801f4cbdee33cdR24) to call spark-submit instead of spark-class directly > the user should be able to configure the resources used by JDBC server > ---------------------------------------------------------------------- > > Key: SPARK-2459 > URL: https://issues.apache.org/jira/browse/SPARK-2459 > Project: Spark > Issue Type: Improvement > Components: SQL > Affects Versions: 1.0.1 > Reporter: Nan Zhu > > I'm trying the jdbc server > I found that the jdbc server always occupies all cores in the cluster > the reason is that when creating HiveContext, it doesn't set anything related > to spark.cores.max or spark.executor.memory > SparkSQLEnv.scala(https://github.com/apache/spark/blob/8032fe2fae3ac40a02c6018c52e76584a14b3438/sql/hive-thriftserver/src/main/scala/org/apache/spark/sql/hive/thriftserver/SparkSQLEnv.scala) > L41-L43 > [~liancheng] -- This message was sent by Atlassian JIRA (v6.2#6252)