Examples with NodeAddedSchedulerEvent org.apache.hadoop.yarn.server.resourcemanager.scheduler.event.NodeAddedSchedulerEvent used on opensource projects

Example 71 with NodeAddedSchedulerEvent

use of org.apache.hadoop.yarn.server.resourcemanager.scheduler.event.NodeAddedSchedulerEvent in project hadoop by apache.

the class TestFairScheduler method testBasicDRFAssignment.

@Test
public void testBasicDRFAssignment() throws Exception {
    scheduler.init(conf);
    scheduler.start();
    scheduler.reinitialize(conf, resourceManager.getRMContext());
    RMNode node = MockNodes.newNodeInfo(1, BuilderUtils.newResource(8192, 5));
    NodeAddedSchedulerEvent nodeEvent = new NodeAddedSchedulerEvent(node);
    scheduler.handle(nodeEvent);
    ApplicationAttemptId appAttId1 = createSchedulingRequest(2048, 1, "queue1", "user1", 2);
    FSAppAttempt app1 = scheduler.getSchedulerApp(appAttId1);
    ApplicationAttemptId appAttId2 = createSchedulingRequest(1024, 2, "queue1", "user1", 2);
    FSAppAttempt app2 = scheduler.getSchedulerApp(appAttId2);
    DominantResourceFairnessPolicy drfPolicy = new DominantResourceFairnessPolicy();
    drfPolicy.initialize(scheduler.getContext());
    scheduler.getQueueManager().getQueue("queue1").setPolicy(drfPolicy);
    scheduler.update();
    // First both apps get a container
    // Then the first gets another container because its dominant share of
    // 2048/8192 is less than the other's of 2/5
    NodeUpdateSchedulerEvent updateEvent = new NodeUpdateSchedulerEvent(node);
    scheduler.handle(updateEvent);
    Assert.assertEquals(1, app1.getLiveContainers().size());
    Assert.assertEquals(0, app2.getLiveContainers().size());
    scheduler.handle(updateEvent);
    Assert.assertEquals(1, app1.getLiveContainers().size());
    Assert.assertEquals(1, app2.getLiveContainers().size());
    scheduler.handle(updateEvent);
    Assert.assertEquals(2, app1.getLiveContainers().size());
    Assert.assertEquals(1, app2.getLiveContainers().size());
}

Example 72 with NodeAddedSchedulerEvent

use of org.apache.hadoop.yarn.server.resourcemanager.scheduler.event.NodeAddedSchedulerEvent in project hadoop by apache.

the class TestFairScheduler method testFixedMaxAssign.

@Test(timeout = 3000)
public void testFixedMaxAssign() throws Exception {
    conf.setBoolean(FairSchedulerConfiguration.ASSIGN_MULTIPLE, true);
    conf.setBoolean(FairSchedulerConfiguration.DYNAMIC_MAX_ASSIGN, false);
    scheduler.init(conf);
    scheduler.start();
    scheduler.reinitialize(conf, resourceManager.getRMContext());
    RMNode node = MockNodes.newNodeInfo(1, Resources.createResource(16384, 16), 0, "127.0.0.1");
    NodeAddedSchedulerEvent nodeEvent = new NodeAddedSchedulerEvent(node);
    NodeUpdateSchedulerEvent updateEvent = new NodeUpdateSchedulerEvent(node);
    scheduler.handle(nodeEvent);
    ApplicationAttemptId attId = createSchedulingRequest(1024, "root.default", "user", 8);
    FSAppAttempt app = scheduler.getSchedulerApp(attId);
    // set maxAssign to 2: only 2 containers should be allocated
    scheduler.maxAssign = 2;
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Incorrect number of containers allocated", 2, app.getLiveContainers().size());
    // set maxAssign to -1: all remaining containers should be allocated
    scheduler.maxAssign = -1;
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Incorrect number of containers allocated", 8, app.getLiveContainers().size());
}

Example 73 with NodeAddedSchedulerEvent

use of org.apache.hadoop.yarn.server.resourcemanager.scheduler.event.NodeAddedSchedulerEvent in project hadoop by apache.

the class TestFifoScheduler method testUpdateResourceOnNode.

@Test(timeout = 2000)
public void testUpdateResourceOnNode() throws Exception {
    AsyncDispatcher dispatcher = new InlineDispatcher();
    Configuration conf = new Configuration();
    RMContainerTokenSecretManager containerTokenSecretManager = new RMContainerTokenSecretManager(conf);
    containerTokenSecretManager.rollMasterKey();
    NMTokenSecretManagerInRM nmTokenSecretManager = new NMTokenSecretManagerInRM(conf);
    nmTokenSecretManager.rollMasterKey();
    RMApplicationHistoryWriter writer = mock(RMApplicationHistoryWriter.class);
    FifoScheduler scheduler = new FifoScheduler();
    RMContext rmContext = new RMContextImpl(dispatcher, null, null, null, null, null, containerTokenSecretManager, nmTokenSecretManager, null, scheduler);
    rmContext.setSystemMetricsPublisher(mock(SystemMetricsPublisher.class));
    rmContext.setRMApplicationHistoryWriter(mock(RMApplicationHistoryWriter.class));
    ((RMContextImpl) rmContext).setYarnConfiguration(new YarnConfiguration());
    NullRMNodeLabelsManager nlm = new NullRMNodeLabelsManager();
    nlm.init(new Configuration());
    rmContext.setNodeLabelManager(nlm);
    scheduler.setRMContext(rmContext);
    ((RMContextImpl) rmContext).setScheduler(scheduler);
    scheduler.init(conf);
    scheduler.start();
    scheduler.reinitialize(new Configuration(), rmContext);
    RMNode node0 = MockNodes.newNodeInfo(1, Resources.createResource(2048, 4), 1, "127.0.0.1");
    NodeAddedSchedulerEvent nodeEvent1 = new NodeAddedSchedulerEvent(node0);
    scheduler.handle(nodeEvent1);
    assertEquals(scheduler.getNumClusterNodes(), 1);
    Resource newResource = Resources.createResource(1024, 4);
    NodeResourceUpdateSchedulerEvent node0ResourceUpdate = new NodeResourceUpdateSchedulerEvent(node0, ResourceOption.newInstance(newResource, ResourceOption.OVER_COMMIT_TIMEOUT_MILLIS_DEFAULT));
    scheduler.handle(node0ResourceUpdate);
    // SchedulerNode's total resource and available resource are changed.
    assertEquals(1024, scheduler.getNodeTracker().getNode(node0.getNodeID()).getTotalResource().getMemorySize());
    assertEquals(1024, scheduler.getNodeTracker().getNode(node0.getNodeID()).getUnallocatedResource().getMemorySize(), 1024);
    QueueInfo queueInfo = scheduler.getQueueInfo(null, false, false);
    Assert.assertEquals(0.0f, queueInfo.getCurrentCapacity(), 0.0f);
    int _appId = 1;
    int _appAttemptId = 1;
    ApplicationAttemptId appAttemptId = createAppAttemptId(_appId, _appAttemptId);
    createMockRMApp(appAttemptId, rmContext);
    AppAddedSchedulerEvent appEvent = new AppAddedSchedulerEvent(appAttemptId.getApplicationId(), "queue1", "user1");
    scheduler.handle(appEvent);
    AppAttemptAddedSchedulerEvent attemptEvent = new AppAttemptAddedSchedulerEvent(appAttemptId, false);
    scheduler.handle(attemptEvent);
    int memory = 1024;
    int priority = 1;
    List<ResourceRequest> ask = new ArrayList<ResourceRequest>();
    ResourceRequest nodeLocal = createResourceRequest(memory, node0.getHostName(), priority, 1);
    ResourceRequest rackLocal = createResourceRequest(memory, node0.getRackName(), priority, 1);
    ResourceRequest any = createResourceRequest(memory, ResourceRequest.ANY, priority, 1);
    ask.add(nodeLocal);
    ask.add(rackLocal);
    ask.add(any);
    scheduler.allocate(appAttemptId, ask, new ArrayList<ContainerId>(), null, null, NULL_UPDATE_REQUESTS);
    // Before the node update event, there are one local request
    Assert.assertEquals(1, nodeLocal.getNumContainers());
    NodeUpdateSchedulerEvent node0Update = new NodeUpdateSchedulerEvent(node0);
    // Now schedule.
    scheduler.handle(node0Update);
    // After the node update event, check no local request
    Assert.assertEquals(0, nodeLocal.getNumContainers());
    // Also check that one container was scheduled
    SchedulerAppReport info = scheduler.getSchedulerAppInfo(appAttemptId);
    Assert.assertEquals(1, info.getLiveContainers().size());
    // And check the default Queue now is full.
    queueInfo = scheduler.getQueueInfo(null, false, false);
    Assert.assertEquals(1.0f, queueInfo.getCurrentCapacity(), 0.0f);
}

Example 74 with NodeAddedSchedulerEvent

use of org.apache.hadoop.yarn.server.resourcemanager.scheduler.event.NodeAddedSchedulerEvent in project hadoop by apache.

the class TestFifoScheduler method testNodeUpdateBeforeAppAttemptInit.

@Test(timeout = 60000)
public void testNodeUpdateBeforeAppAttemptInit() throws Exception {
    FifoScheduler scheduler = new FifoScheduler();
    MockRM rm = new MockRM(conf);
    scheduler.setRMContext(rm.getRMContext());
    scheduler.init(conf);
    scheduler.start();
    scheduler.reinitialize(conf, rm.getRMContext());
    RMNode node = MockNodes.newNodeInfo(1, Resources.createResource(1024, 4), 1, "127.0.0.1");
    scheduler.handle(new NodeAddedSchedulerEvent(node));
    ApplicationId appId = ApplicationId.newInstance(0, 1);
    scheduler.addApplication(appId, "queue1", "user1", false);
    NodeUpdateSchedulerEvent updateEvent = new NodeUpdateSchedulerEvent(node);
    try {
        scheduler.handle(updateEvent);
    } catch (NullPointerException e) {
        Assert.fail();
    }
    ApplicationAttemptId attId = ApplicationAttemptId.newInstance(appId, 1);
    scheduler.addApplicationAttempt(attId, false, false);
    rm.stop();
}

Example 75 with NodeAddedSchedulerEvent

use of org.apache.hadoop.yarn.server.resourcemanager.scheduler.event.NodeAddedSchedulerEvent in project hadoop by apache.

the class TestFairScheduler method testQueueMaxAMShare.

@Test
public void testQueueMaxAMShare() throws Exception {
    conf.set(FairSchedulerConfiguration.ALLOCATION_FILE, ALLOC_FILE);
    PrintWriter out = new PrintWriter(new FileWriter(ALLOC_FILE));
    out.println("<?xml version=\"1.0\"?>");
    out.println("<allocations>");
    out.println("<queue name=\"queue1\">");
    out.println("<maxAMShare>0.2</maxAMShare>");
    out.println("</queue>");
    out.println("</allocations>");
    out.close();
    scheduler.init(conf);
    scheduler.start();
    scheduler.reinitialize(conf, resourceManager.getRMContext());
    RMNode node = MockNodes.newNodeInfo(1, Resources.createResource(20480, 20), 0, "127.0.0.1");
    NodeAddedSchedulerEvent nodeEvent = new NodeAddedSchedulerEvent(node);
    NodeUpdateSchedulerEvent updateEvent = new NodeUpdateSchedulerEvent(node);
    scheduler.handle(nodeEvent);
    scheduler.update();
    FSLeafQueue queue1 = scheduler.getQueueManager().getLeafQueue("queue1", true);
    assertEquals("Queue queue1's fair share should be 0", 0, queue1.getFairShare().getMemorySize());
    createSchedulingRequest(1 * 1024, "root.default", "user1");
    scheduler.update();
    scheduler.handle(updateEvent);
    Resource amResource1 = Resource.newInstance(1024, 1);
    Resource amResource2 = Resource.newInstance(2048, 2);
    Resource amResource3 = Resource.newInstance(1860, 2);
    int amPriority = RMAppAttemptImpl.AM_CONTAINER_PRIORITY.getPriority();
    // Exceeds no limits
    ApplicationAttemptId attId1 = createAppAttemptId(1, 1);
    createApplicationWithAMResource(attId1, "queue1", "user1", amResource1);
    createSchedulingRequestExistingApplication(1024, 1, amPriority, attId1);
    FSAppAttempt app1 = scheduler.getSchedulerApp(attId1);
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Application1's AM requests 1024 MB memory", 1024, app1.getAMResource().getMemorySize());
    assertEquals("Application1's AM should be running", 1, app1.getLiveContainers().size());
    assertEquals("Queue1's AM resource usage should be 1024 MB memory", 1024, queue1.getAmResourceUsage().getMemorySize());
    // Exceeds no limits
    ApplicationAttemptId attId2 = createAppAttemptId(2, 1);
    createApplicationWithAMResource(attId2, "queue1", "user1", amResource1);
    createSchedulingRequestExistingApplication(1024, 1, amPriority, attId2);
    FSAppAttempt app2 = scheduler.getSchedulerApp(attId2);
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Application2's AM requests 1024 MB memory", 1024, app2.getAMResource().getMemorySize());
    assertEquals("Application2's AM should be running", 1, app2.getLiveContainers().size());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // Exceeds queue limit
    ApplicationAttemptId attId3 = createAppAttemptId(3, 1);
    createApplicationWithAMResource(attId3, "queue1", "user1", amResource1);
    createSchedulingRequestExistingApplication(1024, 1, amPriority, attId3);
    FSAppAttempt app3 = scheduler.getSchedulerApp(attId3);
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Application3's AM resource shouldn't be updated", 0, app3.getAMResource().getMemorySize());
    assertEquals("Application3's AM should not be running", 0, app3.getLiveContainers().size());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // Still can run non-AM container
    createSchedulingRequestExistingApplication(1024, 1, attId1);
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Application1 should have two running containers", 2, app1.getLiveContainers().size());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // Remove app1, app3's AM should become running
    AppAttemptRemovedSchedulerEvent appRemovedEvent1 = new AppAttemptRemovedSchedulerEvent(attId1, RMAppAttemptState.FINISHED, false);
    scheduler.update();
    scheduler.handle(appRemovedEvent1);
    scheduler.handle(updateEvent);
    assertEquals("Application1's AM should be finished", 0, app1.getLiveContainers().size());
    assertEquals("Finished application usage should be none", Resources.none(), app1.getResourceUsage());
    assertEquals("Application3's AM should be running", 1, app3.getLiveContainers().size());
    assertEquals("Application3's AM requests 1024 MB memory", 1024, app3.getAMResource().getMemorySize());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // Exceeds queue limit
    ApplicationAttemptId attId4 = createAppAttemptId(4, 1);
    createApplicationWithAMResource(attId4, "queue1", "user1", amResource2);
    createSchedulingRequestExistingApplication(2048, 2, amPriority, attId4);
    FSAppAttempt app4 = scheduler.getSchedulerApp(attId4);
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Application4's AM resource shouldn't be updated", 0, app4.getAMResource().getMemorySize());
    assertEquals("Application4's AM should not be running", 0, app4.getLiveContainers().size());
    assertEquals("Finished application usage should be none", Resources.none(), app4.getResourceUsage());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // Exceeds queue limit
    ApplicationAttemptId attId5 = createAppAttemptId(5, 1);
    createApplicationWithAMResource(attId5, "queue1", "user1", amResource2);
    createSchedulingRequestExistingApplication(2048, 2, amPriority, attId5);
    FSAppAttempt app5 = scheduler.getSchedulerApp(attId5);
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Application5's AM resource shouldn't be updated", 0, app5.getAMResource().getMemorySize());
    assertEquals("Application5's AM should not be running", 0, app5.getLiveContainers().size());
    assertEquals("Finished application usage should be none", Resources.none(), app5.getResourceUsage());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // Remove un-running app doesn't affect others
    AppAttemptRemovedSchedulerEvent appRemovedEvent4 = new AppAttemptRemovedSchedulerEvent(attId4, RMAppAttemptState.KILLED, false);
    scheduler.handle(appRemovedEvent4);
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Application5's AM should not be running", 0, app5.getLiveContainers().size());
    assertEquals("Finished application usage should be none", Resources.none(), app5.getResourceUsage());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // Remove app2 and app3, app5's AM should become running
    AppAttemptRemovedSchedulerEvent appRemovedEvent2 = new AppAttemptRemovedSchedulerEvent(attId2, RMAppAttemptState.FINISHED, false);
    AppAttemptRemovedSchedulerEvent appRemovedEvent3 = new AppAttemptRemovedSchedulerEvent(attId3, RMAppAttemptState.FINISHED, false);
    scheduler.handle(appRemovedEvent2);
    scheduler.handle(appRemovedEvent3);
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Application2's AM should be finished", 0, app2.getLiveContainers().size());
    assertEquals("Finished application usage should be none", Resources.none(), app2.getResourceUsage());
    assertEquals("Application3's AM should be finished", 0, app3.getLiveContainers().size());
    assertEquals("Finished application usage should be none", Resources.none(), app3.getResourceUsage());
    assertEquals("Application5's AM should be running", 1, app5.getLiveContainers().size());
    assertEquals("Application5's AM requests 2048 MB memory", 2048, app5.getAMResource().getMemorySize());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // request non-AM container for app5
    createSchedulingRequestExistingApplication(1024, 1, attId5);
    assertEquals("Application5's AM should have 1 container", 1, app5.getLiveContainers().size());
    // complete AM container before non-AM container is allocated.
    // spark application hit this situation.
    RMContainer amContainer5 = (RMContainer) app5.getLiveContainers().toArray()[0];
    ContainerExpiredSchedulerEvent containerExpired = new ContainerExpiredSchedulerEvent(amContainer5.getContainerId());
    scheduler.handle(containerExpired);
    assertEquals("Application5's AM should have 0 container", 0, app5.getLiveContainers().size());
    assertEquals("Finished application usage should be none", Resources.none(), app5.getResourceUsage());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    scheduler.update();
    scheduler.handle(updateEvent);
    // non-AM container should be allocated
    // check non-AM container allocation is not rejected
    // due to queue MaxAMShare limitation.
    assertEquals("Application5 should have 1 container", 1, app5.getLiveContainers().size());
    // check non-AM container allocation won't affect queue AmResourceUsage
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // Check amResource normalization
    ApplicationAttemptId attId6 = createAppAttemptId(6, 1);
    createApplicationWithAMResource(attId6, "queue1", "user1", amResource3);
    createSchedulingRequestExistingApplication(1860, 2, amPriority, attId6);
    FSAppAttempt app6 = scheduler.getSchedulerApp(attId6);
    scheduler.update();
    scheduler.handle(updateEvent);
    assertEquals("Application6's AM should not be running", 0, app6.getLiveContainers().size());
    assertEquals("Finished application usage should be none", Resources.none(), app6.getResourceUsage());
    assertEquals("Application6's AM resource shouldn't be updated", 0, app6.getAMResource().getMemorySize());
    assertEquals("Queue1's AM resource usage should be 2048 MB memory", 2048, queue1.getAmResourceUsage().getMemorySize());
    // Remove all apps
    AppAttemptRemovedSchedulerEvent appRemovedEvent5 = new AppAttemptRemovedSchedulerEvent(attId5, RMAppAttemptState.FINISHED, false);
    AppAttemptRemovedSchedulerEvent appRemovedEvent6 = new AppAttemptRemovedSchedulerEvent(attId6, RMAppAttemptState.FINISHED, false);
    scheduler.handle(appRemovedEvent5);
    scheduler.handle(appRemovedEvent6);
    scheduler.update();
    assertEquals("Queue1's AM resource usage should be 0", 0, queue1.getAmResourceUsage().getMemorySize());
}