Search in sources :

Example 1 with DefaultAllocatedSlotPool

use of org.apache.flink.runtime.jobmaster.slotpool.DefaultAllocatedSlotPool in project flink by apache.

the class AdaptiveSchedulerTest method testNumRestartsMetric.

@Test
public void testNumRestartsMetric() throws Exception {
    final CompletableFuture<Gauge<Long>> numRestartsMetricFuture = new CompletableFuture<>();
    final MetricRegistry metricRegistry = TestingMetricRegistry.builder().setRegisterConsumer((metric, name, group) -> {
        if (MetricNames.NUM_RESTARTS.equals(name)) {
            numRestartsMetricFuture.complete((Gauge<Long>) metric);
        }
    }).build();
    final JobGraph jobGraph = createJobGraph();
    final DefaultDeclarativeSlotPool declarativeSlotPool = new DefaultDeclarativeSlotPool(jobGraph.getJobID(), new DefaultAllocatedSlotPool(), ignored -> {
    }, Time.minutes(10), Time.minutes(10));
    final Configuration configuration = new Configuration();
    configuration.set(JobManagerOptions.MIN_PARALLELISM_INCREASE, 1);
    configuration.set(JobManagerOptions.RESOURCE_WAIT_TIMEOUT, Duration.ofMillis(1L));
    final AdaptiveScheduler scheduler = new AdaptiveSchedulerBuilder(jobGraph, singleThreadMainThreadExecutor).setJobMasterConfiguration(configuration).setJobManagerJobMetricGroup(JobManagerMetricGroup.createJobManagerMetricGroup(metricRegistry, "localhost").addJob(new JobID(), "jobName")).setDeclarativeSlotPool(declarativeSlotPool).build();
    final Gauge<Long> numRestartsMetric = numRestartsMetricFuture.get();
    final SubmissionBufferingTaskManagerGateway taskManagerGateway = new SubmissionBufferingTaskManagerGateway(1 + PARALLELISM);
    taskManagerGateway.setCancelConsumer(createCancelConsumer(scheduler));
    singleThreadMainThreadExecutor.execute(() -> {
        scheduler.startScheduling();
        declarativeSlotPool.offerSlots(createSlotOffersForResourceRequirements(ResourceCounter.withResource(ResourceProfile.UNKNOWN, 1)), new LocalTaskManagerLocation(), taskManagerGateway, System.currentTimeMillis());
    });
    // wait for the first task submission
    taskManagerGateway.waitForSubmissions(1, Duration.ofSeconds(5));
    assertThat(numRestartsMetric.getValue()).isEqualTo(0L);
    singleThreadMainThreadExecutor.execute(() -> {
        // offer more slots, which will cause a restart in order to scale up
        offerSlots(declarativeSlotPool, createSlotOffersForResourceRequirements(ResourceCounter.withResource(ResourceProfile.UNKNOWN, PARALLELISM)), taskManagerGateway);
    });
    // wait for the second task submissions
    taskManagerGateway.waitForSubmissions(PARALLELISM, Duration.ofSeconds(5));
    assertThat(numRestartsMetric.getValue()).isEqualTo(1L);
}
Also used : Arrays(java.util.Arrays) ResourceRequirement(org.apache.flink.runtime.slots.ResourceRequirement) TaskNotRunningException(org.apache.flink.runtime.operators.coordination.TaskNotRunningException) ArchivedExecution(org.apache.flink.runtime.executiongraph.ArchivedExecution) TestingSlotAllocator(org.apache.flink.runtime.scheduler.adaptive.allocator.TestingSlotAllocator) CheckpointException(org.apache.flink.runtime.checkpoint.CheckpointException) ResultPartitionID(org.apache.flink.runtime.io.network.partition.ResultPartitionID) TestingFatalErrorHandler(org.apache.flink.runtime.util.TestingFatalErrorHandler) MetricRegistry(org.apache.flink.runtime.metrics.MetricRegistry) Duration(java.time.Duration) ClassRule(org.junit.ClassRule) TestingCheckpointRecoveryFactory(org.apache.flink.runtime.checkpoint.TestingCheckpointRecoveryFactory) ManuallyTriggeredComponentMainThreadExecutor(org.apache.flink.runtime.concurrent.ManuallyTriggeredComponentMainThreadExecutor) BlockingQueue(java.util.concurrent.BlockingQueue) MetricOptions(org.apache.flink.configuration.MetricOptions) JobManagerOptions(org.apache.flink.configuration.JobManagerOptions) Executors(java.util.concurrent.Executors) MetricNames(org.apache.flink.runtime.metrics.MetricNames) ArrayBlockingQueue(java.util.concurrent.ArrayBlockingQueue) VertexParallelismStore(org.apache.flink.runtime.scheduler.VertexParallelismStore) Time(org.apache.flink.api.common.time.Time) RootExceptionHistoryEntry(org.apache.flink.runtime.scheduler.exceptionhistory.RootExceptionHistoryEntry) FlinkException(org.apache.flink.util.FlinkException) ComponentMainThreadExecutor(org.apache.flink.runtime.concurrent.ComponentMainThreadExecutor) UpTimeGauge(org.apache.flink.runtime.executiongraph.metrics.UpTimeGauge) ExceptionHistoryEntry(org.apache.flink.runtime.scheduler.exceptionhistory.ExceptionHistoryEntry) LocalTaskManagerLocation(org.apache.flink.runtime.taskmanager.LocalTaskManagerLocation) ResourceCounter(org.apache.flink.runtime.util.ResourceCounter) JobStatus(org.apache.flink.api.common.JobStatus) DefaultAllocatedSlotPool(org.apache.flink.runtime.jobmaster.slotpool.DefaultAllocatedSlotPool) ArrayList(java.util.ArrayList) SchedulerNG(org.apache.flink.runtime.scheduler.SchedulerNG) DownTimeGauge(org.apache.flink.runtime.executiongraph.metrics.DownTimeGauge) PartitionProducerDisposedException(org.apache.flink.runtime.jobmanager.PartitionProducerDisposedException) Gauge(org.apache.flink.metrics.Gauge) ScheduledExecutorService(java.util.concurrent.ScheduledExecutorService) BiConsumer(java.util.function.BiConsumer) FixedDelayRestartBackoffTimeStrategy(org.apache.flink.runtime.executiongraph.failover.flip1.FixedDelayRestartBackoffTimeStrategy) Nullable(javax.annotation.Nullable) ArchivedExecutionVertex(org.apache.flink.runtime.executiongraph.ArchivedExecutionVertex) TestExecutorResource(org.apache.flink.testutils.executor.TestExecutorResource) ExecutionState(org.apache.flink.runtime.execution.ExecutionState) CheckpointsCleaner(org.apache.flink.runtime.checkpoint.CheckpointsCleaner) TestOperatorEvent(org.apache.flink.runtime.operators.coordination.TestOperatorEvent) Test(org.junit.Test) IOException(java.io.IOException) IterableUtils(org.apache.flink.util.IterableUtils) SimpleAckingTaskManagerGateway(org.apache.flink.runtime.executiongraph.utils.SimpleAckingTaskManagerGateway) ExecutionException(java.util.concurrent.ExecutionException) TaskExecutionStateTransition(org.apache.flink.runtime.executiongraph.TaskExecutionStateTransition) JobID(org.apache.flink.api.common.JobID) NoRestartBackoffTimeStrategy(org.apache.flink.runtime.executiongraph.failover.flip1.NoRestartBackoffTimeStrategy) SlotPoolTestUtils.offerSlots(org.apache.flink.runtime.jobmaster.slotpool.SlotPoolTestUtils.offerSlots) JobManagerMetricGroup(org.apache.flink.runtime.metrics.groups.JobManagerMetricGroup) ComponentMainThreadExecutorServiceAdapter(org.apache.flink.runtime.concurrent.ComponentMainThreadExecutorServiceAdapter) Assertions.assertThat(org.assertj.core.api.Assertions.assertThat) JobGraph(org.apache.flink.runtime.jobgraph.JobGraph) CheckpointCoordinatorConfiguration(org.apache.flink.runtime.jobgraph.tasks.CheckpointCoordinatorConfiguration) SuppressRestartsException(org.apache.flink.runtime.execution.SuppressRestartsException) ArchivedExecutionGraphTest(org.apache.flink.runtime.executiongraph.ArchivedExecutionGraphTest) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) TestLogger(org.apache.flink.util.TestLogger) JobCheckpointingSettings(org.apache.flink.runtime.jobgraph.tasks.JobCheckpointingSettings) CheckpointIDCounter(org.apache.flink.runtime.checkpoint.CheckpointIDCounter) ArchivedExecutionJobVertex(org.apache.flink.runtime.executiongraph.ArchivedExecutionJobVertex) TestingCompletedCheckpointStore(org.apache.flink.runtime.checkpoint.TestingCompletedCheckpointStore) TestRestartBackoffTimeStrategy(org.apache.flink.runtime.executiongraph.failover.flip1.TestRestartBackoffTimeStrategy) IntermediateDataSetID(org.apache.flink.runtime.jobgraph.IntermediateDataSetID) Collectors(java.util.stream.Collectors) ResourceProfile(org.apache.flink.runtime.clusterframework.types.ResourceProfile) DefaultDeclarativeSlotPool(org.apache.flink.runtime.jobmaster.slotpool.DefaultDeclarativeSlotPool) List(java.util.List) StandaloneCheckpointIDCounter(org.apache.flink.runtime.checkpoint.StandaloneCheckpointIDCounter) CoordinationRequest(org.apache.flink.runtime.operators.coordination.CoordinationRequest) OperatorID(org.apache.flink.runtime.jobgraph.OperatorID) Optional(java.util.Optional) SchedulerExecutionMode(org.apache.flink.configuration.SchedulerExecutionMode) KeyGroupRangeAssignment(org.apache.flink.runtime.state.KeyGroupRangeAssignment) OneShotLatch(org.apache.flink.core.testutils.OneShotLatch) SchedulerBase(org.apache.flink.runtime.scheduler.SchedulerBase) SavepointFormatType(org.apache.flink.core.execution.SavepointFormatType) JobVertex(org.apache.flink.runtime.jobgraph.JobVertex) AtomicBoolean(java.util.concurrent.atomic.AtomicBoolean) CompletableFuture(java.util.concurrent.CompletableFuture) VertexParallelismInformation(org.apache.flink.runtime.scheduler.VertexParallelismInformation) TaskDeploymentDescriptor(org.apache.flink.runtime.deployment.TaskDeploymentDescriptor) TestingCheckpointIDCounter(org.apache.flink.runtime.checkpoint.TestingCheckpointIDCounter) TestingMetricRegistry(org.apache.flink.runtime.metrics.util.TestingMetricRegistry) Nonnull(javax.annotation.Nonnull) StandaloneCompletedCheckpointStore(org.apache.flink.runtime.checkpoint.StandaloneCompletedCheckpointStore) DefaultSchedulerTest(org.apache.flink.runtime.scheduler.DefaultSchedulerTest) JobGraphTestUtils.streamingJobGraph(org.apache.flink.runtime.jobgraph.JobGraphTestUtils.streamingJobGraph) ArchivedExecutionGraph(org.apache.flink.runtime.executiongraph.ArchivedExecutionGraph) ArchivedExecutionGraphBuilder(org.apache.flink.runtime.rest.handler.legacy.utils.ArchivedExecutionGraphBuilder) Logger(org.slf4j.Logger) Configuration(org.apache.flink.configuration.Configuration) CompletedCheckpointStore(org.apache.flink.runtime.checkpoint.CompletedCheckpointStore) JobStatusListener(org.apache.flink.runtime.executiongraph.JobStatusListener) DefaultDeclarativeSlotPoolTest.createSlotOffersForResourceRequirements(org.apache.flink.runtime.jobmaster.slotpool.DefaultDeclarativeSlotPoolTest.createSlotOffersForResourceRequirements) TimeUnit(java.util.concurrent.TimeUnit) Consumer(java.util.function.Consumer) ExecutionAttemptID(org.apache.flink.runtime.executiongraph.ExecutionAttemptID) ExecutionGraphTestUtils.createNoOpVertex(org.apache.flink.runtime.executiongraph.ExecutionGraphTestUtils.createNoOpVertex) SchedulerTestingUtils.enableCheckpointing(org.apache.flink.runtime.scheduler.SchedulerTestingUtils.enableCheckpointing) TaskExecutionState(org.apache.flink.runtime.taskmanager.TaskExecutionState) TemporaryFolder(org.junit.rules.TemporaryFolder) CheckpointCoordinatorConfiguration(org.apache.flink.runtime.jobgraph.tasks.CheckpointCoordinatorConfiguration) Configuration(org.apache.flink.configuration.Configuration) MetricRegistry(org.apache.flink.runtime.metrics.MetricRegistry) TestingMetricRegistry(org.apache.flink.runtime.metrics.util.TestingMetricRegistry) UpTimeGauge(org.apache.flink.runtime.executiongraph.metrics.UpTimeGauge) DownTimeGauge(org.apache.flink.runtime.executiongraph.metrics.DownTimeGauge) Gauge(org.apache.flink.metrics.Gauge) CompletableFuture(java.util.concurrent.CompletableFuture) JobGraph(org.apache.flink.runtime.jobgraph.JobGraph) JobGraphTestUtils.streamingJobGraph(org.apache.flink.runtime.jobgraph.JobGraphTestUtils.streamingJobGraph) LocalTaskManagerLocation(org.apache.flink.runtime.taskmanager.LocalTaskManagerLocation) DefaultAllocatedSlotPool(org.apache.flink.runtime.jobmaster.slotpool.DefaultAllocatedSlotPool) DefaultDeclarativeSlotPool(org.apache.flink.runtime.jobmaster.slotpool.DefaultDeclarativeSlotPool) JobID(org.apache.flink.api.common.JobID) Test(org.junit.Test) ArchivedExecutionGraphTest(org.apache.flink.runtime.executiongraph.ArchivedExecutionGraphTest) DefaultSchedulerTest(org.apache.flink.runtime.scheduler.DefaultSchedulerTest)

Aggregations

IOException (java.io.IOException)1 Duration (java.time.Duration)1 ArrayList (java.util.ArrayList)1 Arrays (java.util.Arrays)1 List (java.util.List)1 Optional (java.util.Optional)1 ArrayBlockingQueue (java.util.concurrent.ArrayBlockingQueue)1 BlockingQueue (java.util.concurrent.BlockingQueue)1 CompletableFuture (java.util.concurrent.CompletableFuture)1 ExecutionException (java.util.concurrent.ExecutionException)1 Executors (java.util.concurrent.Executors)1 ScheduledExecutorService (java.util.concurrent.ScheduledExecutorService)1 TimeUnit (java.util.concurrent.TimeUnit)1 AtomicBoolean (java.util.concurrent.atomic.AtomicBoolean)1 AtomicInteger (java.util.concurrent.atomic.AtomicInteger)1 BiConsumer (java.util.function.BiConsumer)1 Consumer (java.util.function.Consumer)1 Collectors (java.util.stream.Collectors)1 Nonnull (javax.annotation.Nonnull)1 Nullable (javax.annotation.Nullable)1