0001
0002
0003
0004
0005
0006
0007
0008
0009
0010
0011
0012
0013
0014
0015
0016
0017
0018 package org.apache.spark.examples;
0019
0020 import org.apache.spark.SparkJobInfo;
0021 import org.apache.spark.SparkStageInfo;
0022 import org.apache.spark.api.java.JavaFutureAction;
0023 import org.apache.spark.api.java.JavaRDD;
0024 import org.apache.spark.api.java.JavaSparkContext;
0025 import org.apache.spark.api.java.function.Function;
0026 import org.apache.spark.sql.SparkSession;
0027
0028 import java.util.Arrays;
0029 import java.util.List;
0030
0031
0032
0033
0034 public final class JavaStatusTrackerDemo {
0035
0036 public static final String APP_NAME = "JavaStatusAPIDemo";
0037
0038 public static final class IdentityWithDelay<T> implements Function<T, T> {
0039 @Override
0040 public T call(T x) throws Exception {
0041 Thread.sleep(2 * 1000);
0042 return x;
0043 }
0044 }
0045
0046 public static void main(String[] args) throws Exception {
0047 SparkSession spark = SparkSession
0048 .builder()
0049 .appName(APP_NAME)
0050 .getOrCreate();
0051
0052 JavaSparkContext jsc = new JavaSparkContext(spark.sparkContext());
0053
0054
0055 JavaRDD<Integer> rdd = jsc.parallelize(Arrays.asList(1, 2, 3, 4, 5), 5).map(
0056 new IdentityWithDelay<>());
0057 JavaFutureAction<List<Integer>> jobFuture = rdd.collectAsync();
0058 while (!jobFuture.isDone()) {
0059 Thread.sleep(1000);
0060 List<Integer> jobIds = jobFuture.jobIds();
0061 if (jobIds.isEmpty()) {
0062 continue;
0063 }
0064 int currentJobId = jobIds.get(jobIds.size() - 1);
0065 SparkJobInfo jobInfo = jsc.statusTracker().getJobInfo(currentJobId);
0066 SparkStageInfo stageInfo = jsc.statusTracker().getStageInfo(jobInfo.stageIds()[0]);
0067 System.out.println(stageInfo.numTasks() + " tasks total: " + stageInfo.numActiveTasks() +
0068 " active, " + stageInfo.numCompletedTasks() + " complete");
0069 }
0070
0071 System.out.println("Job results are: " + jobFuture.get());
0072 spark.stop();
0073 }
0074 }