You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@pig.apache.org by xu...@apache.org on 2016/11/29 04:01:46 UTC
svn commit: r1771840 - in
/pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark:
SparkLauncher.java SparkUtil.java
Author: xuefu
Date: Tue Nov 29 04:01:46 2016
New Revision: 1771840
URL: http://svn.apache.org/viewvc?rev=1771840&view=rev
Log:
PIG-5068: Set SPARK_REDUCERS by pig.properties not by system configuration (Liyun via Xuefu)
Modified:
pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkLauncher.java
pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkUtil.java
Modified: pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkLauncher.java
URL: http://svn.apache.org/viewvc/pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkLauncher.java?rev=1771840&r1=1771839&r2=1771840&view=diff
==============================================================================
--- pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkLauncher.java (original)
+++ pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkLauncher.java Tue Nov 29 04:01:46 2016
@@ -650,5 +650,9 @@ public class SparkLauncher extends Launc
SchemaTupleBackend.initialize(jobConf, pigContext);
Utils.setDefaultTimeZone(jobConf);
PigMapReduce.sJobConfInternal.set(jobConf);
+ String sparkReducers = pigContext.getProperties().getProperty("spark.reducers");
+ if (sparkReducers != null) {
+ SparkUtil.setSparkReducers(Integer.parseInt(sparkReducers));
+ }
}
}
Modified: pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkUtil.java
URL: http://svn.apache.org/viewvc/pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkUtil.java?rev=1771840&r1=1771839&r2=1771840&view=diff
==============================================================================
--- pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkUtil.java (original)
+++ pig/branches/spark/src/org/apache/pig/backend/hadoop/executionengine/spark/SparkUtil.java Tue Nov 29 04:01:46 2016
@@ -25,7 +25,6 @@ import java.util.Random;
import java.util.UUID;
import java.util.concurrent.ConcurrentHashMap;
-import org.apache.spark.broadcast.Broadcast;
import scala.Product2;
import scala.Tuple2;
import scala.collection.JavaConversions;
@@ -55,10 +54,12 @@ import org.apache.pig.impl.plan.PlanExce
import org.apache.pig.impl.util.ObjectSerializer;
import org.apache.spark.HashPartitioner;
import org.apache.spark.Partitioner;
+import org.apache.spark.broadcast.Broadcast;
import org.apache.spark.rdd.RDD;
public class SparkUtil {
+ private static ThreadLocal<Integer> SPARK_REDUCERS = null;
private static ConcurrentHashMap<String, Broadcast<List<Tuple>>> broadcastedVars = new ConcurrentHashMap() ;
public static <T> ClassTag<T> getManifest(Class<T> clazz) {
@@ -130,10 +131,8 @@ public class SparkUtil {
public static int getParallelism(List<RDD<Tuple>> predecessors,
PhysicalOperator physicalOperator) {
-
- String numReducers = System.getenv("SPARK_REDUCERS");
- if (numReducers != null) {
- return Integer.parseInt(numReducers);
+ if (SPARK_REDUCERS != null) {
+ return getSparkReducers();
}
int parallelism = physicalOperator.getRequestedParallelism();
@@ -178,4 +177,13 @@ public class SparkUtil {
static public ConcurrentHashMap<String, Broadcast<List<Tuple>>> getBroadcastedVars() {
return broadcastedVars;
}
+
+
+ public static int getSparkReducers() {
+ return SPARK_REDUCERS.get();
+ }
+
+ public static void setSparkReducers(int sparkReducers) {
+ SPARK_REDUCERS.set(sparkReducers);
+ }
}
\ No newline at end of file