Cheng Lian created SPARK-6507: --------------------------------- Summary: Create separate Hive Driver instance for each SQL query in HiveThriftServer2 Key: SPARK-6507 URL: https://issues.apache.org/jira/browse/SPARK-6507 Project: Spark Issue Type: Bug Components: SQL Affects Versions: 1.3.0, 1.2.1, 1.1.1, 1.0.2 Reporter: Cheng Lian
In the current implementation of HiveThriftServer2, Hive {{Driver}} instances are cached and reused among queries. However, {{Driver}} is not thread-safe, and may cause racing conditions. In SPARK-4908, we synchronized {{HiveContext.runHive}} to avoid this issue, but this affects concurrency negatively, because no two native commands can be executed concurrently. This is pretty bad for heavy commands like ANALYZE. Please refer [this comment|https://issues.apache.org/jira/browse/SPARK-4908?focusedCommentId=14264469&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14264469] in SPARK-4908 for details. -- This message was sent by Atlassian JIRA (v6.3.4#6332) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org