Use SQLConf to configure in-memory columnar caching

marmbrus · marmbrus · commit 23620825bee3 · 2014-07-29T18:19:51.000-07:00
diff --git a/sql/core/src/main/scala/org/apache/spark/sql/SQLConf.scala b/sql/core/src/main/scala/org/apache/spark/sql/SQLConf.scala
@@ -22,6 +22,7 @@ import java.util.Properties
 import scala.collection.JavaConverters._
 
 object SQLConf {
+  val COMPRESS_CACHED = "spark.sql.inMemoryColumnarStorage.compressed"
   val AUTO_BROADCASTJOIN_THRESHOLD = "spark.sql.autoBroadcastJoinThreshold"
   val SHUFFLE_PARTITIONS = "spark.sql.shuffle.partitions"
   val DEFAULT_SIZE_IN_BYTES = "spark.sql.defaultSizeInBytes"
@@ -49,6 +50,9 @@ trait SQLConf {
   /** ************************ Spark SQL Params/Hints ******************* */
   // TODO: refactor so that these hints accessors don't pollute the name space of SQLContext?
 
+  /** When true tables cached using the in-memory columnar caching will be compressed. */
+  private[spark] def useCompression: Boolean = get(COMPRESS_CACHED, "false").toBoolean
+
   /** Number of partitions to use for shuffle operators. */
   private[spark] def numShufflePartitions: Int = get(SHUFFLE_PARTITIONS, "200").toInt
 
diff --git a/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala b/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala
@@ -192,8 +192,6 @@ class SQLContext(@transient val sparkContext: SparkContext)
         currentTable.logicalPlan
 
       case _ =>
-        val useCompression =
-          sparkContext.conf.getBoolean("spark.sql.inMemoryColumnarStorage.compressed", false)
         InMemoryRelation(useCompression, executePlan(currentTable).executedPlan)
     }
 

Original file line number	Diff line number	Diff line change
`@@ -192,8 +192,6 @@ class SQLContext(@transient val sparkContext: SparkContext)`
`192`	`192`	`currentTable.logicalPlan`
`193`	`193`
`194`	`194`	`case _ =>`
`195`		`- val useCompression =`
`196`		`- sparkContext.conf.getBoolean("spark.sql.inMemoryColumnarStorage.compressed", false)`
`197`	`195`	`InMemoryRelation(useCompression, executePlan(currentTable).executedPlan)`
`198`	`196`	`}`
`199`	`197`