MAPREDUCE-6454. Fixed MapReduce to modify HADOOP_CLASSPATH to have distributed cache files so that child processes running hadoop scripts can access these files. Contributed by Junping Du.

2015-08-20 17:35:24 -07:00 · 2015-08-20 17:35:24 -07:00 · 9da81a3955
parent 45be1240d9
commit 9da81a3955
5 changed files with 98 additions and 30 deletions
--- a/hadoop-mapreduce-project/CHANGES.txt
+++ b/hadoop-mapreduce-project/CHANGES.txt
@ -557,6 +557,10 @@ Release 2.6.2 - UNRELEASED

  BUG FIXES

+    MAPREDUCE-6454. Fixed MapReduce to modify HADOOP_CLASSPATH to have distributed
+    cache files so that child processes running hadoop scripts can access these
+    files. (Junping Du via vinodkv)
+
 Release 2.6.1 - UNRELEASED

  INCOMPATIBLE CHANGES
--- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/job/impl/TaskAttemptImpl.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/job/impl/TaskAttemptImpl.java
@ -173,6 +173,7 @@ public abstract class TaskAttemptImpl implements
  private static AtomicBoolean initialClasspathFlag = new AtomicBoolean();
  private static String initialClasspath = null;
  private static String initialAppClasspath = null;
+  private static String initialHadoopClasspath = null;
  private static Object commonContainerSpecLock = new Object();
  private static ContainerLaunchContext commonContainerSpec = null;
  private static final Object classpathLock = new Object();
@ -742,6 +743,7 @@ public abstract class TaskAttemptImpl implements
      MRApps.setClasspath(env, conf);
      initialClasspath = env.get(Environment.CLASSPATH.name());
      initialAppClasspath = env.get(Environment.APP_CLASSPATH.name());
+      initialHadoopClasspath = env.get(Environment.HADOOP_CLASSPATH.name());
      initialClasspathFlag.set(true);
      return initialClasspath;
    }
@ -879,6 +881,13 @@ public abstract class TaskAttemptImpl implements
          Environment.CLASSPATH.name(),
          getInitialClasspath(conf), conf);

+      if (initialHadoopClasspath != null) {
+        MRApps.addToEnvironment(
+            environment,
+            Environment.HADOOP_CLASSPATH.name(),
+            initialHadoopClasspath, conf);
+      }
+
      if (initialAppClasspath != null) {
        MRApps.addToEnvironment(
            environment,  
--- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/main/java/org/apache/hadoop/mapreduce/v2/util/MRApps.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/main/java/org/apache/hadoop/mapreduce/v2/util/MRApps.java
@ -245,11 +245,30 @@ public class MRApps extends Apps {
      conf.getBoolean(MRJobConfig.MAPREDUCE_JOB_CLASSLOADER, false)
        ? Environment.APP_CLASSPATH.name() : Environment.CLASSPATH.name();

+    String hadoopClasspathEnvVar = Environment.HADOOP_CLASSPATH.name();
+
    MRApps.addToEnvironment(environment,
      classpathEnvVar, crossPlatformifyMREnv(conf, Environment.PWD), conf);
+
+    MRApps.addToEnvironment(environment,
+        hadoopClasspathEnvVar, crossPlatformifyMREnv(conf, Environment.PWD),
+        conf);
+
    if (!userClassesTakesPrecedence) {
      MRApps.setMRFrameworkClasspath(environment, conf);
    }
+
+    addClasspathToEnv(environment, classpathEnvVar, conf);
+    addClasspathToEnv(environment, hadoopClasspathEnvVar, conf);
+
+    if (userClassesTakesPrecedence) {
+      MRApps.setMRFrameworkClasspath(environment, conf);
+    }
+  }
+
+  @SuppressWarnings("deprecation")
+  public static void addClasspathToEnv(Map<String, String> environment,
+      String classpathEnvVar, Configuration conf) throws IOException {
    MRApps.addToEnvironment(
        environment,
        classpathEnvVar,
@ -257,15 +276,21 @@ public class MRApps extends Apps {
    MRApps.addToEnvironment(
        environment,
        classpathEnvVar,
-        MRJobConfig.JOB_JAR + Path.SEPARATOR + "classes" + Path.SEPARATOR, conf);
+        MRJobConfig.JOB_JAR + Path.SEPARATOR + "classes" + Path.SEPARATOR,
+        conf);
+
    MRApps.addToEnvironment(
        environment,
        classpathEnvVar,
-        MRJobConfig.JOB_JAR + Path.SEPARATOR + "lib" + Path.SEPARATOR + "*", conf);
+        MRJobConfig.JOB_JAR + Path.SEPARATOR + "lib" + Path.SEPARATOR + "*",
+        conf);
+
    MRApps.addToEnvironment(
        environment,
        classpathEnvVar,
-        crossPlatformifyMREnv(conf, Environment.PWD) + Path.SEPARATOR + "*", conf);
+        crossPlatformifyMREnv(conf, Environment.PWD) + Path.SEPARATOR + "*",
+        conf);
+
    // a * in the classpath will only find a .jar, so we need to filter out
    // all .jars and add everything else
    addToClasspathIfNotJar(DistributedCache.getFileClassPaths(conf),
@ -276,9 +301,6 @@ public class MRApps extends Apps {
        DistributedCache.getCacheArchives(conf),
        conf,
        environment, classpathEnvVar);
-    if (userClassesTakesPrecedence) {
-      MRApps.setMRFrameworkClasspath(environment, conf);
-    }
  }

  /**
@ -446,6 +468,7 @@ public class MRApps extends Apps {
    return startCommitFile;
  }

+  @SuppressWarnings("deprecation")
  public static void setupDistributedCache( 
      Configuration conf, 
      Map<String, LocalResource> localResources) 
@ -478,6 +501,7 @@ public class MRApps extends Apps {
   * @param conf
   * @throws java.io.IOException
   */
+  @SuppressWarnings("deprecation")
  public static void setupDistributedCacheLocal(Configuration conf)
      throws IOException {

@ -545,6 +569,7 @@ public class MRApps extends Apps {
  // TODO - Move this to MR!
  // Use TaskDistributedCacheManager.CacheFiles.makeCacheFiles(URI[], 
  // long[], boolean[], Path[], FileType)
+  @SuppressWarnings("deprecation")
  private static void parseDistributedCacheArtifacts(
      Configuration conf,
      Map<String, LocalResource> localResources,
--- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapreduce/v2/util/TestMRApps.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapreduce/v2/util/TestMRApps.java
@ -36,6 +36,8 @@ import java.util.HashMap;
 import java.util.List;
 import java.util.Map;

+import org.apache.commons.logging.Log;
+import org.apache.commons.logging.LogFactory;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.FileSystem;
 import org.apache.hadoop.fs.FilterFileSystem;
@ -67,6 +69,9 @@ import org.junit.Test;
 public class TestMRApps {
  private static File testWorkDir = null;

+  private static final Log LOG =
+      LogFactory.getLog(TestMRApps.class);
+
  @BeforeClass
  public static void setupTestDirs() throws IOException {
    testWorkDir = new File("target", TestMRApps.class.getCanonicalName());
@ -193,8 +198,8 @@ public class TestMRApps {
    Job job = Job.getInstance(conf);
    Map<String, String> environment = new HashMap<String, String>();
    MRApps.setClasspath(environment, job.getConfiguration());
-    assertTrue(environment.get("CLASSPATH").startsWith(
-      ApplicationConstants.Environment.PWD.$$()
+    assertTrue(environment.get(ApplicationConstants.Environment.CLASSPATH.
+        name()).startsWith(ApplicationConstants.Environment.PWD.$$()
            + ApplicationConstants.CLASS_PATH_SEPARATOR));
    String yarnAppClasspath = job.getConfiguration().get(
        YarnConfiguration.YARN_APPLICATION_CLASSPATH,
@ -205,7 +210,8 @@ public class TestMRApps {
          yarnAppClasspath.replaceAll(",\\s*",
            ApplicationConstants.CLASS_PATH_SEPARATOR).trim();
    }
-    assertTrue(environment.get("CLASSPATH").contains(yarnAppClasspath));
+    assertTrue(environment.get(ApplicationConstants.Environment.
+        CLASSPATH.name()).contains(yarnAppClasspath));
    String mrAppClasspath =
        job.getConfiguration().get(
            MRJobConfig.MAPREDUCE_APPLICATION_CLASSPATH,
@ -215,7 +221,8 @@ public class TestMRApps {
          mrAppClasspath.replaceAll(",\\s*",
            ApplicationConstants.CLASS_PATH_SEPARATOR).trim();
    }
-    assertTrue(environment.get("CLASSPATH").contains(mrAppClasspath));
+    assertTrue(environment.get(ApplicationConstants.Environment.CLASSPATH.
+        name()).contains(mrAppClasspath));
  }
  
  @Test (timeout = 120000)
@ -234,8 +241,12 @@ public class TestMRApps {
    conf.set(MRJobConfig.CACHE_ARCHIVES, testTGZQualifiedPath + "#testTGZ");
    Map<String, String> environment = new HashMap<String, String>();
    MRApps.setClasspath(environment, conf);
-    assertTrue(environment.get("CLASSPATH").startsWith(
+    assertTrue(environment.get(ApplicationConstants.Environment.CLASSPATH.name()).startsWith(
      ApplicationConstants.Environment.PWD.$$() + ApplicationConstants.CLASS_PATH_SEPARATOR));
+    assertTrue(environment.get(ApplicationConstants.Environment.
+        HADOOP_CLASSPATH.name()).startsWith(
+            ApplicationConstants.Environment.PWD.$$() +
+            ApplicationConstants.CLASS_PATH_SEPARATOR));
    String confClasspath = job.getConfiguration().get(
        YarnConfiguration.YARN_APPLICATION_CLASSPATH,
        StringUtils.join(",",
@ -244,8 +255,19 @@ public class TestMRApps {
      confClasspath = confClasspath.replaceAll(",\\s*", ApplicationConstants.CLASS_PATH_SEPARATOR)
        .trim();
    }
-    assertTrue(environment.get("CLASSPATH").contains(confClasspath));
-    assertTrue(environment.get("CLASSPATH").contains("testTGZ"));
+    LOG.info("CLASSPATH: "+ environment.get(
+        ApplicationConstants.Environment.CLASSPATH.name()));
+    LOG.info("confClasspath: " + confClasspath);
+    assertTrue(environment.get(
+        ApplicationConstants.Environment.CLASSPATH.name()).contains(
+            confClasspath));
+    LOG.info("HADOOP_CLASSPATH: " + environment.get(
+        ApplicationConstants.Environment.HADOOP_CLASSPATH.name()));
+    assertTrue(environment.get(
+        ApplicationConstants.Environment.CLASSPATH.name()).contains("testTGZ"));
+    assertTrue(environment.get(
+        ApplicationConstants.Environment.HADOOP_CLASSPATH.name()).
+        contains("testTGZ"));
  }

 @Test (timeout = 120000)
@ -259,7 +281,7 @@ public class TestMRApps {
    } catch (Exception e) {
      fail("Got exception while setting classpath");
    }
-    String env_str = env.get("CLASSPATH");
+    String env_str = env.get(ApplicationConstants.Environment.CLASSPATH.name());
    String expectedClasspath = StringUtils.join(ApplicationConstants.CLASS_PATH_SEPARATOR,
      Arrays.asList(ApplicationConstants.Environment.PWD.$$(), "job.jar/job.jar",
        "job.jar/classes/", "job.jar/lib/*",
@ -279,7 +301,7 @@ public class TestMRApps {
    } catch (Exception e) {
      fail("Got exception while setting classpath");
    }
-    String env_str = env.get("CLASSPATH");
+    String env_str = env.get(ApplicationConstants.Environment.CLASSPATH.name());
    String expectedClasspath = StringUtils.join(ApplicationConstants.CLASS_PATH_SEPARATOR,
      Arrays.asList("job.jar/job.jar", "job.jar/classes/", "job.jar/lib/*",
        ApplicationConstants.Environment.PWD.$$() + "/*"));
@ -296,8 +318,9 @@ public class TestMRApps {
    conf.setBoolean(MRJobConfig.MAPREDUCE_JOB_CLASSLOADER, true);
    Map<String, String> env = new HashMap<String, String>();
    MRApps.setClasspath(env, conf);
-    String cp = env.get("CLASSPATH");
-    String appCp = env.get("APP_CLASSPATH");
+    String cp = env.get(ApplicationConstants.Environment.CLASSPATH.name());
+    String appCp = env.get(ApplicationConstants.Environment.
+        APP_CLASSPATH.name());
    assertFalse("MAPREDUCE_JOB_CLASSLOADER true, but job.jar is in the"
      + " classpath!", cp.contains("jar" + ApplicationConstants.CLASS_PATH_SEPARATOR + "job"));
    assertFalse("MAPREDUCE_JOB_CLASSLOADER true, but PWD is in the classpath!",
@ -338,7 +361,8 @@ public class TestMRApps {
        Arrays.asList(ApplicationConstants.Environment.PWD.$$(),
            FRAMEWORK_CLASSPATH, stdClasspath));
    assertEquals("Incorrect classpath with framework and no user precedence",
-        expectedClasspath, env.get("CLASSPATH"));
+        expectedClasspath, env.get(ApplicationConstants.Environment.
+            CLASSPATH.name()));

    env.clear();
    conf.setBoolean(MRJobConfig.MAPREDUCE_JOB_USER_CLASSPATH_FIRST, true);
@ -347,7 +371,8 @@ public class TestMRApps {
        Arrays.asList(ApplicationConstants.Environment.PWD.$$(),
            stdClasspath, FRAMEWORK_CLASSPATH));
    assertEquals("Incorrect classpath with framework and user precedence",
-        expectedClasspath, env.get("CLASSPATH"));
+        expectedClasspath, env.get(ApplicationConstants.Environment.CLASSPATH.
+            name()));
  }

  @Test (timeout = 30000)
--- a/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-api/src/main/java/org/apache/hadoop/yarn/api/ApplicationConstants.java
+++ b/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-api/src/main/java/org/apache/hadoop/yarn/api/ApplicationConstants.java
@ -160,6 +160,11 @@ public interface ApplicationConstants {
     */
    APP_CLASSPATH("APP_CLASSPATH"),
    
+    /**
+     * $HADOOP_CLASSPATH.
+     */
+    HADOOP_CLASSPATH("HADOOP_CLASSPATH"),
+
    /**
     * $LD_LIBRARY_PATH
     */