Search in sources :

Example 21 with AppContext

use of org.apache.hadoop.mapreduce.v2.app.AppContext in project hadoop by apache.

the class TestContainerLauncher method testPoolSize.

@Test(timeout = 10000)
public void testPoolSize() throws InterruptedException {
    ApplicationId appId = ApplicationId.newInstance(12345, 67);
    ApplicationAttemptId appAttemptId = ApplicationAttemptId.newInstance(appId, 3);
    JobId jobId = MRBuilderUtils.newJobId(appId, 8);
    TaskId taskId = MRBuilderUtils.newTaskId(jobId, 9, TaskType.MAP);
    AppContext context = mock(AppContext.class);
    CustomContainerLauncher containerLauncher = new CustomContainerLauncher(context);
    containerLauncher.init(new Configuration());
    containerLauncher.start();
    ThreadPoolExecutor threadPool = containerLauncher.getThreadPool();
    // No events yet
    Assert.assertEquals(containerLauncher.initialPoolSize, MRJobConfig.DEFAULT_MR_AM_CONTAINERLAUNCHER_THREADPOOL_INITIAL_SIZE);
    Assert.assertEquals(0, threadPool.getPoolSize());
    Assert.assertEquals(containerLauncher.initialPoolSize, threadPool.getCorePoolSize());
    Assert.assertNull(containerLauncher.foundErrors);
    containerLauncher.expectedCorePoolSize = containerLauncher.initialPoolSize;
    for (int i = 0; i < 10; i++) {
        ContainerId containerId = ContainerId.newContainerId(appAttemptId, i);
        TaskAttemptId taskAttemptId = MRBuilderUtils.newTaskAttemptId(taskId, i);
        containerLauncher.handle(new ContainerLauncherEvent(taskAttemptId, containerId, "host" + i + ":1234", null, ContainerLauncher.EventType.CONTAINER_REMOTE_LAUNCH));
    }
    waitForEvents(containerLauncher, 10);
    Assert.assertEquals(10, threadPool.getPoolSize());
    Assert.assertNull(containerLauncher.foundErrors);
    // Same set of hosts, so no change
    containerLauncher.finishEventHandling = true;
    int timeOut = 0;
    while (containerLauncher.numEventsProcessed.get() < 10 && timeOut++ < 200) {
        LOG.info("Waiting for number of events processed to become " + 10 + ". It is now " + containerLauncher.numEventsProcessed.get() + ". Timeout is " + timeOut);
        Thread.sleep(1000);
    }
    Assert.assertEquals(10, containerLauncher.numEventsProcessed.get());
    containerLauncher.finishEventHandling = false;
    for (int i = 0; i < 10; i++) {
        ContainerId containerId = ContainerId.newContainerId(appAttemptId, i + 10);
        TaskAttemptId taskAttemptId = MRBuilderUtils.newTaskAttemptId(taskId, i + 10);
        containerLauncher.handle(new ContainerLauncherEvent(taskAttemptId, containerId, "host" + i + ":1234", null, ContainerLauncher.EventType.CONTAINER_REMOTE_LAUNCH));
    }
    waitForEvents(containerLauncher, 20);
    Assert.assertEquals(10, threadPool.getPoolSize());
    Assert.assertNull(containerLauncher.foundErrors);
    // Different hosts, there should be an increase in core-thread-pool size to
    // 21(11hosts+10buffer)
    // Core pool size should be 21 but the live pool size should be only 11.
    containerLauncher.expectedCorePoolSize = 11 + containerLauncher.initialPoolSize;
    containerLauncher.finishEventHandling = false;
    ContainerId containerId = ContainerId.newContainerId(appAttemptId, 21);
    TaskAttemptId taskAttemptId = MRBuilderUtils.newTaskAttemptId(taskId, 21);
    containerLauncher.handle(new ContainerLauncherEvent(taskAttemptId, containerId, "host11:1234", null, ContainerLauncher.EventType.CONTAINER_REMOTE_LAUNCH));
    waitForEvents(containerLauncher, 21);
    Assert.assertEquals(11, threadPool.getPoolSize());
    Assert.assertNull(containerLauncher.foundErrors);
    containerLauncher.stop();
    // change configuration MR_AM_CONTAINERLAUNCHER_THREADPOOL_INITIAL_SIZE
    // and verify initialPoolSize value.
    Configuration conf = new Configuration();
    conf.setInt(MRJobConfig.MR_AM_CONTAINERLAUNCHER_THREADPOOL_INITIAL_SIZE, 20);
    containerLauncher = new CustomContainerLauncher(context);
    containerLauncher.init(conf);
    Assert.assertEquals(containerLauncher.initialPoolSize, 20);
}
Also used : TaskId(org.apache.hadoop.mapreduce.v2.api.records.TaskId) Configuration(org.apache.hadoop.conf.Configuration) YarnConfiguration(org.apache.hadoop.yarn.conf.YarnConfiguration) ContainerId(org.apache.hadoop.yarn.api.records.ContainerId) TaskAttemptId(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId) AppContext(org.apache.hadoop.mapreduce.v2.app.AppContext) ApplicationAttemptId(org.apache.hadoop.yarn.api.records.ApplicationAttemptId) ThreadPoolExecutor(java.util.concurrent.ThreadPoolExecutor) ApplicationId(org.apache.hadoop.yarn.api.records.ApplicationId) JobId(org.apache.hadoop.mapreduce.v2.api.records.JobId) Test(org.junit.Test)

Example 22 with AppContext

use of org.apache.hadoop.mapreduce.v2.app.AppContext in project hadoop by apache.

the class TestRMContainerAllocator method testUnregistrationOnlyIfRegistered.

@Test
public void testUnregistrationOnlyIfRegistered() throws Exception {
    Configuration conf = new Configuration();
    final MyResourceManager rm = new MyResourceManager(conf);
    rm.start();
    DrainDispatcher rmDispatcher = (DrainDispatcher) rm.getRMContext().getDispatcher();
    // Submit the application
    RMApp rmApp = rm.submitApp(1024);
    rmDispatcher.await();
    MockNM amNodeManager = rm.registerNode("127.0.0.1:1234", 11264);
    amNodeManager.nodeHeartbeat(true);
    rmDispatcher.await();
    final ApplicationAttemptId appAttemptId = rmApp.getCurrentAppAttempt().getAppAttemptId();
    rm.sendAMLaunched(appAttemptId);
    rmDispatcher.await();
    MRApp mrApp = new MRApp(appAttemptId, ContainerId.newContainerId(appAttemptId, 0), 10, 0, false, this.getClass().getName(), true, 1) {

        @Override
        protected Dispatcher createDispatcher() {
            return new DrainDispatcher();
        }

        protected ContainerAllocator createContainerAllocator(ClientService clientService, AppContext context) {
            return new MyContainerAllocator(rm, appAttemptId, context);
        }

        ;
    };
    mrApp.submit(conf);
    DrainDispatcher amDispatcher = (DrainDispatcher) mrApp.getDispatcher();
    MyContainerAllocator allocator = (MyContainerAllocator) mrApp.getContainerAllocator();
    amDispatcher.await();
    Assert.assertTrue(allocator.isApplicationMasterRegistered());
    mrApp.stop();
    Assert.assertTrue(allocator.isUnregistered());
}
Also used : DrainDispatcher(org.apache.hadoop.yarn.event.DrainDispatcher) RMApp(org.apache.hadoop.yarn.server.resourcemanager.rmapp.RMApp) Configuration(org.apache.hadoop.conf.Configuration) YarnConfiguration(org.apache.hadoop.yarn.conf.YarnConfiguration) ClientService(org.apache.hadoop.mapreduce.v2.app.client.ClientService) MockNM(org.apache.hadoop.yarn.server.resourcemanager.MockNM) RunningAppContext(org.apache.hadoop.mapreduce.v2.app.MRAppMaster.RunningAppContext) AppContext(org.apache.hadoop.mapreduce.v2.app.AppContext) ApplicationAttemptId(org.apache.hadoop.yarn.api.records.ApplicationAttemptId) MRApp(org.apache.hadoop.mapreduce.v2.app.MRApp) Test(org.junit.Test)

Example 23 with AppContext

use of org.apache.hadoop.mapreduce.v2.app.AppContext in project hadoop by apache.

the class TestTaskAttempt method testAppDiognosticEventOnNewTask.

@Test
public void testAppDiognosticEventOnNewTask() throws Exception {
    ApplicationId appId = ApplicationId.newInstance(1, 2);
    ApplicationAttemptId appAttemptId = ApplicationAttemptId.newInstance(appId, 0);
    JobId jobId = MRBuilderUtils.newJobId(appId, 1);
    TaskId taskId = MRBuilderUtils.newTaskId(jobId, 1, TaskType.MAP);
    TaskAttemptId attemptId = MRBuilderUtils.newTaskAttemptId(taskId, 0);
    Path jobFile = mock(Path.class);
    MockEventHandler eventHandler = new MockEventHandler();
    TaskAttemptListener taListener = mock(TaskAttemptListener.class);
    when(taListener.getAddress()).thenReturn(new InetSocketAddress("localhost", 0));
    JobConf jobConf = new JobConf();
    jobConf.setClass("fs.file.impl", StubbedFS.class, FileSystem.class);
    jobConf.setBoolean("fs.file.impl.disable.cache", true);
    jobConf.set(JobConf.MAPRED_MAP_TASK_ENV, "");
    jobConf.set(MRJobConfig.APPLICATION_ATTEMPT_ID, "10");
    TaskSplitMetaInfo splits = mock(TaskSplitMetaInfo.class);
    when(splits.getLocations()).thenReturn(new String[] { "127.0.0.1" });
    AppContext appCtx = mock(AppContext.class);
    ClusterInfo clusterInfo = mock(ClusterInfo.class);
    Resource resource = mock(Resource.class);
    when(appCtx.getClusterInfo()).thenReturn(clusterInfo);
    when(resource.getMemorySize()).thenReturn(1024L);
    setupTaskAttemptFinishingMonitor(eventHandler, jobConf, appCtx);
    TaskAttemptImpl taImpl = new MapTaskAttemptImpl(taskId, 1, eventHandler, jobFile, 1, splits, jobConf, taListener, new Token(), new Credentials(), SystemClock.getInstance(), appCtx);
    NodeId nid = NodeId.newInstance("127.0.0.1", 0);
    ContainerId contId = ContainerId.newContainerId(appAttemptId, 3);
    Container container = mock(Container.class);
    when(container.getId()).thenReturn(contId);
    when(container.getNodeId()).thenReturn(nid);
    when(container.getNodeHttpAddress()).thenReturn("localhost:0");
    taImpl.handle(new TaskAttemptDiagnosticsUpdateEvent(attemptId, "Task got killed"));
    assertFalse("InternalError occurred trying to handle TA_DIAGNOSTICS_UPDATE on assigned task", eventHandler.internalError);
}
Also used : Path(org.apache.hadoop.fs.Path) TaskId(org.apache.hadoop.mapreduce.v2.api.records.TaskId) TaskAttemptId(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId) TaskAttemptListener(org.apache.hadoop.mapreduce.v2.app.TaskAttemptListener) InetSocketAddress(java.net.InetSocketAddress) TaskAttemptDiagnosticsUpdateEvent(org.apache.hadoop.mapreduce.v2.app.job.event.TaskAttemptDiagnosticsUpdateEvent) AppContext(org.apache.hadoop.mapreduce.v2.app.AppContext) Resource(org.apache.hadoop.yarn.api.records.Resource) MapTaskAttemptImpl(org.apache.hadoop.mapred.MapTaskAttemptImpl) Token(org.apache.hadoop.security.token.Token) ApplicationAttemptId(org.apache.hadoop.yarn.api.records.ApplicationAttemptId) ClusterInfo(org.apache.hadoop.mapreduce.v2.app.ClusterInfo) Container(org.apache.hadoop.yarn.api.records.Container) ContainerId(org.apache.hadoop.yarn.api.records.ContainerId) NodeId(org.apache.hadoop.yarn.api.records.NodeId) TaskSplitMetaInfo(org.apache.hadoop.mapreduce.split.JobSplit.TaskSplitMetaInfo) ApplicationId(org.apache.hadoop.yarn.api.records.ApplicationId) MapTaskAttemptImpl(org.apache.hadoop.mapred.MapTaskAttemptImpl) JobConf(org.apache.hadoop.mapred.JobConf) JobId(org.apache.hadoop.mapreduce.v2.api.records.JobId) Credentials(org.apache.hadoop.security.Credentials) Test(org.junit.Test)

Example 24 with AppContext

use of org.apache.hadoop.mapreduce.v2.app.AppContext in project hadoop by apache.

the class TestTaskAttempt method testContainerKillWhileRunning.

@Test
public void testContainerKillWhileRunning() throws Exception {
    ApplicationId appId = ApplicationId.newInstance(1, 2);
    ApplicationAttemptId appAttemptId = ApplicationAttemptId.newInstance(appId, 0);
    JobId jobId = MRBuilderUtils.newJobId(appId, 1);
    TaskId taskId = MRBuilderUtils.newTaskId(jobId, 1, TaskType.MAP);
    TaskAttemptId attemptId = MRBuilderUtils.newTaskAttemptId(taskId, 0);
    Path jobFile = mock(Path.class);
    MockEventHandler eventHandler = new MockEventHandler();
    TaskAttemptListener taListener = mock(TaskAttemptListener.class);
    when(taListener.getAddress()).thenReturn(new InetSocketAddress("localhost", 0));
    JobConf jobConf = new JobConf();
    jobConf.setClass("fs.file.impl", StubbedFS.class, FileSystem.class);
    jobConf.setBoolean("fs.file.impl.disable.cache", true);
    jobConf.set(JobConf.MAPRED_MAP_TASK_ENV, "");
    jobConf.set(MRJobConfig.APPLICATION_ATTEMPT_ID, "10");
    TaskSplitMetaInfo splits = mock(TaskSplitMetaInfo.class);
    when(splits.getLocations()).thenReturn(new String[] { "127.0.0.1" });
    AppContext appCtx = mock(AppContext.class);
    ClusterInfo clusterInfo = mock(ClusterInfo.class);
    Resource resource = mock(Resource.class);
    when(appCtx.getClusterInfo()).thenReturn(clusterInfo);
    when(resource.getMemorySize()).thenReturn(1024L);
    TaskAttemptImpl taImpl = new MapTaskAttemptImpl(taskId, 1, eventHandler, jobFile, 1, splits, jobConf, taListener, new Token(), new Credentials(), SystemClock.getInstance(), appCtx);
    NodeId nid = NodeId.newInstance("127.0.0.2", 0);
    ContainerId contId = ContainerId.newContainerId(appAttemptId, 3);
    Container container = mock(Container.class);
    when(container.getId()).thenReturn(contId);
    when(container.getNodeId()).thenReturn(nid);
    when(container.getNodeHttpAddress()).thenReturn("localhost:0");
    taImpl.handle(new TaskAttemptEvent(attemptId, TaskAttemptEventType.TA_SCHEDULE));
    taImpl.handle(new TaskAttemptContainerAssignedEvent(attemptId, container, mock(Map.class)));
    taImpl.handle(new TaskAttemptContainerLaunchedEvent(attemptId, 0));
    assertEquals("Task attempt is not in running state", taImpl.getState(), TaskAttemptState.RUNNING);
    taImpl.handle(new TaskAttemptEvent(attemptId, TaskAttemptEventType.TA_KILL));
    assertFalse("InternalError occurred trying to handle TA_KILL", eventHandler.internalError);
    assertEquals("Task should be in KILL_CONTAINER_CLEANUP state", TaskAttemptStateInternal.KILL_CONTAINER_CLEANUP, taImpl.getInternalState());
}
Also used : Path(org.apache.hadoop.fs.Path) TaskId(org.apache.hadoop.mapreduce.v2.api.records.TaskId) TaskAttemptId(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId) TaskAttemptListener(org.apache.hadoop.mapreduce.v2.app.TaskAttemptListener) InetSocketAddress(java.net.InetSocketAddress) AppContext(org.apache.hadoop.mapreduce.v2.app.AppContext) Resource(org.apache.hadoop.yarn.api.records.Resource) MapTaskAttemptImpl(org.apache.hadoop.mapred.MapTaskAttemptImpl) Token(org.apache.hadoop.security.token.Token) TaskAttemptEvent(org.apache.hadoop.mapreduce.v2.app.job.event.TaskAttemptEvent) ApplicationAttemptId(org.apache.hadoop.yarn.api.records.ApplicationAttemptId) TaskAttemptContainerAssignedEvent(org.apache.hadoop.mapreduce.v2.app.job.event.TaskAttemptContainerAssignedEvent) TaskAttemptContainerLaunchedEvent(org.apache.hadoop.mapreduce.v2.app.job.event.TaskAttemptContainerLaunchedEvent) ClusterInfo(org.apache.hadoop.mapreduce.v2.app.ClusterInfo) Container(org.apache.hadoop.yarn.api.records.Container) ContainerId(org.apache.hadoop.yarn.api.records.ContainerId) NodeId(org.apache.hadoop.yarn.api.records.NodeId) TaskSplitMetaInfo(org.apache.hadoop.mapreduce.split.JobSplit.TaskSplitMetaInfo) ApplicationId(org.apache.hadoop.yarn.api.records.ApplicationId) MapTaskAttemptImpl(org.apache.hadoop.mapred.MapTaskAttemptImpl) JobConf(org.apache.hadoop.mapred.JobConf) JobId(org.apache.hadoop.mapreduce.v2.api.records.JobId) Credentials(org.apache.hadoop.security.Credentials) Test(org.junit.Test)

Example 25 with AppContext

use of org.apache.hadoop.mapreduce.v2.app.AppContext in project hadoop by apache.

the class TestTaskAttempt method testContainerCleanedWhileRunning.

@Test
public void testContainerCleanedWhileRunning() throws Exception {
    ApplicationId appId = ApplicationId.newInstance(1, 2);
    ApplicationAttemptId appAttemptId = ApplicationAttemptId.newInstance(appId, 0);
    JobId jobId = MRBuilderUtils.newJobId(appId, 1);
    TaskId taskId = MRBuilderUtils.newTaskId(jobId, 1, TaskType.MAP);
    TaskAttemptId attemptId = MRBuilderUtils.newTaskAttemptId(taskId, 0);
    Path jobFile = mock(Path.class);
    MockEventHandler eventHandler = new MockEventHandler();
    TaskAttemptListener taListener = mock(TaskAttemptListener.class);
    when(taListener.getAddress()).thenReturn(new InetSocketAddress("localhost", 0));
    JobConf jobConf = new JobConf();
    jobConf.setClass("fs.file.impl", StubbedFS.class, FileSystem.class);
    jobConf.setBoolean("fs.file.impl.disable.cache", true);
    jobConf.set(JobConf.MAPRED_MAP_TASK_ENV, "");
    jobConf.set(MRJobConfig.APPLICATION_ATTEMPT_ID, "10");
    TaskSplitMetaInfo splits = mock(TaskSplitMetaInfo.class);
    when(splits.getLocations()).thenReturn(new String[] { "127.0.0.1" });
    AppContext appCtx = mock(AppContext.class);
    ClusterInfo clusterInfo = mock(ClusterInfo.class);
    Resource resource = mock(Resource.class);
    when(appCtx.getClusterInfo()).thenReturn(clusterInfo);
    when(resource.getMemorySize()).thenReturn(1024L);
    setupTaskAttemptFinishingMonitor(eventHandler, jobConf, appCtx);
    TaskAttemptImpl taImpl = new MapTaskAttemptImpl(taskId, 1, eventHandler, jobFile, 1, splits, jobConf, taListener, new Token(), new Credentials(), SystemClock.getInstance(), appCtx);
    NodeId nid = NodeId.newInstance("127.0.0.2", 0);
    ContainerId contId = ContainerId.newContainerId(appAttemptId, 3);
    Container container = mock(Container.class);
    when(container.getId()).thenReturn(contId);
    when(container.getNodeId()).thenReturn(nid);
    when(container.getNodeHttpAddress()).thenReturn("localhost:0");
    taImpl.handle(new TaskAttemptEvent(attemptId, TaskAttemptEventType.TA_SCHEDULE));
    taImpl.handle(new TaskAttemptContainerAssignedEvent(attemptId, container, mock(Map.class)));
    taImpl.handle(new TaskAttemptContainerLaunchedEvent(attemptId, 0));
    assertEquals("Task attempt is not in running state", taImpl.getState(), TaskAttemptState.RUNNING);
    taImpl.handle(new TaskAttemptEvent(attemptId, TaskAttemptEventType.TA_CONTAINER_CLEANED));
    assertFalse("InternalError occurred trying to handle TA_CONTAINER_CLEANED", eventHandler.internalError);
    assertEquals("Task attempt is not assigned on the local rack", Locality.RACK_LOCAL, taImpl.getLocality());
}
Also used : Path(org.apache.hadoop.fs.Path) TaskId(org.apache.hadoop.mapreduce.v2.api.records.TaskId) TaskAttemptId(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId) TaskAttemptListener(org.apache.hadoop.mapreduce.v2.app.TaskAttemptListener) InetSocketAddress(java.net.InetSocketAddress) AppContext(org.apache.hadoop.mapreduce.v2.app.AppContext) Resource(org.apache.hadoop.yarn.api.records.Resource) MapTaskAttemptImpl(org.apache.hadoop.mapred.MapTaskAttemptImpl) Token(org.apache.hadoop.security.token.Token) TaskAttemptEvent(org.apache.hadoop.mapreduce.v2.app.job.event.TaskAttemptEvent) ApplicationAttemptId(org.apache.hadoop.yarn.api.records.ApplicationAttemptId) TaskAttemptContainerAssignedEvent(org.apache.hadoop.mapreduce.v2.app.job.event.TaskAttemptContainerAssignedEvent) TaskAttemptContainerLaunchedEvent(org.apache.hadoop.mapreduce.v2.app.job.event.TaskAttemptContainerLaunchedEvent) ClusterInfo(org.apache.hadoop.mapreduce.v2.app.ClusterInfo) Container(org.apache.hadoop.yarn.api.records.Container) ContainerId(org.apache.hadoop.yarn.api.records.ContainerId) NodeId(org.apache.hadoop.yarn.api.records.NodeId) TaskSplitMetaInfo(org.apache.hadoop.mapreduce.split.JobSplit.TaskSplitMetaInfo) ApplicationId(org.apache.hadoop.yarn.api.records.ApplicationId) MapTaskAttemptImpl(org.apache.hadoop.mapred.MapTaskAttemptImpl) JobConf(org.apache.hadoop.mapred.JobConf) JobId(org.apache.hadoop.mapreduce.v2.api.records.JobId) Credentials(org.apache.hadoop.security.Credentials) Test(org.junit.Test)

Aggregations

Test (org.junit.Test)74 AppContext (org.apache.hadoop.mapreduce.v2.app.AppContext)73 JobId (org.apache.hadoop.mapreduce.v2.api.records.JobId)47 TaskId (org.apache.hadoop.mapreduce.v2.api.records.TaskId)32 Job (org.apache.hadoop.mapreduce.v2.app.job.Job)32 Configuration (org.apache.hadoop.conf.Configuration)31 TaskAttemptId (org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId)26 ApplicationAttemptId (org.apache.hadoop.yarn.api.records.ApplicationAttemptId)22 ContainerId (org.apache.hadoop.yarn.api.records.ContainerId)21 Path (org.apache.hadoop.fs.Path)18 MockAppContext (org.apache.hadoop.mapreduce.v2.app.MockAppContext)18 ApplicationId (org.apache.hadoop.yarn.api.records.ApplicationId)18 Task (org.apache.hadoop.mapreduce.v2.app.job.Task)17 Container (org.apache.hadoop.yarn.api.records.Container)14 JobConf (org.apache.hadoop.mapred.JobConf)13 TaskSplitMetaInfo (org.apache.hadoop.mapreduce.split.JobSplit.TaskSplitMetaInfo)13 InetSocketAddress (java.net.InetSocketAddress)12 ClusterInfo (org.apache.hadoop.mapreduce.v2.app.ClusterInfo)12 TaskAttemptListener (org.apache.hadoop.mapreduce.v2.app.TaskAttemptListener)12 TaskAttemptEvent (org.apache.hadoop.mapreduce.v2.app.job.event.TaskAttemptEvent)12