GitHub user rajeshbalamohan opened a pull request: https://github.com/apache/spark/pull/10846
SPARK-12920. [SQL]. Spark thrift server can run at very high CPU with⦠Spark thrift server runs at very high CPU when concurrent users submit queries to the system over a period of time. Profiling revealed it is due to many Conf objects getting cacheed in HadoopRDD and this causing lots of pressure for GC when running queries on datasets with large number of partitions. Also, job UI retention causes issues with large jobs. Details are mentioned in SPARK-12920 and profiler snapshots are attached. Fix introduces "spark.hadoop.cacheConf" to optionally cache the jobConf in HadoopRDD. JobProgressListener fixes are related to trimming the job/stage details. You can merge this pull request into a Git repository by running: $ git pull https://github.com/rajeshbalamohan/spark SPARK-12920 Alternatively you can review and apply these changes as the patch at: https://github.com/apache/spark/pull/10846.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #10846 ---- commit 12d77beb0c6c4e3f336aab43a5907775da715753 Author: Rajesh Balamohan <rbalamo...@apache.org> Date: 2016-01-20T09:14:36Z SPARK-12920. [SQL]. Spark thrift server can run at very high CPU with concurrent users ---- --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org