Examples with CancellationSuccess org.apache.flink.runtime.messages.JobManagerMessages.CancellationSuccess used on opensource projects

Example 6 with CancellationSuccess

use of org.apache.flink.runtime.messages.JobManagerMessages.CancellationSuccess in project flink by apache.

the class CliFrontend method cancel.

/**
	 * Executes the CANCEL action.
	 * 
	 * @param args Command line arguments for the cancel action.
	 */
protected int cancel(String[] args) {
    LOG.info("Running 'cancel' command.");
    CancelOptions options;
    try {
        options = CliFrontendParser.parseCancelCommand(args);
    } catch (CliArgsException e) {
        return handleArgException(e);
    } catch (Throwable t) {
        return handleError(t);
    }
    // evaluate help flag
    if (options.isPrintHelp()) {
        CliFrontendParser.printHelpForCancel();
        return 0;
    }
    String[] cleanedArgs = options.getArgs();
    boolean withSavepoint = options.isWithSavepoint();
    String targetDirectory = options.getSavepointTargetDirectory();
    JobID jobId;
    // - cancel -s <targetDir> <jobID> => custom target dir (parsed correctly)
    if (cleanedArgs.length > 0) {
        String jobIdString = cleanedArgs[0];
        try {
            jobId = new JobID(StringUtils.hexStringToByte(jobIdString));
        } catch (Exception e) {
            LOG.error("Error: The value for the Job ID is not a valid ID.");
            System.out.println("Error: The value for the Job ID is not a valid ID.");
            return 1;
        }
    } else if (targetDirectory != null) {
        // Try this for case: cancel -s <jobID> (default savepoint target dir)
        String jobIdString = targetDirectory;
        try {
            jobId = new JobID(StringUtils.hexStringToByte(jobIdString));
            targetDirectory = null;
        } catch (Exception e) {
            LOG.error("Missing JobID in the command line arguments.");
            System.out.println("Error: Specify a Job ID to cancel a job.");
            return 1;
        }
    } else {
        LOG.error("Missing JobID in the command line arguments.");
        System.out.println("Error: Specify a Job ID to cancel a job.");
        return 1;
    }
    try {
        ActorGateway jobManager = getJobManagerGateway(options);
        Object cancelMsg;
        if (withSavepoint) {
            if (targetDirectory == null) {
                logAndSysout("Cancelling job " + jobId + " with savepoint to default savepoint directory.");
            } else {
                logAndSysout("Cancelling job " + jobId + " with savepoint to " + targetDirectory + ".");
            }
            cancelMsg = new CancelJobWithSavepoint(jobId, targetDirectory);
        } else {
            logAndSysout("Cancelling job " + jobId + ".");
            cancelMsg = new CancelJob(jobId);
        }
        Future<Object> response = jobManager.ask(cancelMsg, clientTimeout);
        final Object rc = Await.result(response, clientTimeout);
        if (rc instanceof CancellationSuccess) {
            if (withSavepoint) {
                CancellationSuccess success = (CancellationSuccess) rc;
                String savepointPath = success.savepointPath();
                logAndSysout("Cancelled job " + jobId + ". Savepoint stored in " + savepointPath + ".");
            } else {
                logAndSysout("Cancelled job " + jobId + ".");
            }
        } else if (rc instanceof CancellationFailure) {
            throw new Exception("Canceling the job with ID " + jobId + " failed.", ((CancellationFailure) rc).cause());
        } else {
            throw new IllegalStateException("Unexpected response: " + rc);
        }
        return 0;
    } catch (Throwable t) {
        return handleError(t);
    }
}

Example 7 with CancellationSuccess

use of org.apache.flink.runtime.messages.JobManagerMessages.CancellationSuccess in project flink by apache.

the class JobCancellationWithSavepointHandlersTest method testTriggerNewRequest.

/**
	 * Tests triggering a new request and monitoring it.
	 */
@Test
public void testTriggerNewRequest() throws Exception {
    JobID jobId = new JobID();
    ExecutionGraphHolder holder = mock(ExecutionGraphHolder.class);
    ExecutionGraph graph = mock(ExecutionGraph.class);
    CheckpointCoordinator coord = mock(CheckpointCoordinator.class);
    when(holder.getExecutionGraph(eq(jobId), any(ActorGateway.class))).thenReturn(graph);
    when(graph.getCheckpointCoordinator()).thenReturn(coord);
    JobCancellationWithSavepointHandlers handlers = new JobCancellationWithSavepointHandlers(holder, EC);
    JobCancellationWithSavepointHandlers.TriggerHandler trigger = handlers.getTriggerHandler();
    JobCancellationWithSavepointHandlers.InProgressHandler progress = handlers.getInProgressHandler();
    Map<String, String> params = new HashMap<>();
    params.put("jobid", jobId.toString());
    params.put("targetDirectory", "custom-directory");
    ActorGateway jobManager = mock(ActorGateway.class);
    // Successful
    Promise<Object> promise = new Promise.DefaultPromise<>();
    when(jobManager.ask(any(Object.class), any(FiniteDuration.class))).thenReturn(promise);
    // Trigger
    FullHttpResponse response = trigger.handleRequest(params, Collections.<String, String>emptyMap(), jobManager);
    verify(jobManager).ask(eq(new CancelJobWithSavepoint(jobId, "custom-directory")), any(FiniteDuration.class));
    String location = String.format("/jobs/%s/cancel-with-savepoint/in-progress/1", jobId);
    assertEquals(HttpResponseStatus.ACCEPTED, response.getStatus());
    assertEquals("application/json", response.headers().get(HttpHeaders.Names.CONTENT_TYPE));
    assertEquals(Integer.toString(response.content().readableBytes()), response.headers().get(HttpHeaders.Names.CONTENT_LENGTH));
    assertEquals(location, response.headers().get(HttpHeaders.Names.LOCATION));
    String json = response.content().toString(Charset.forName("UTF-8"));
    JsonNode root = new ObjectMapper().readTree(json);
    assertEquals("accepted", root.get("status").getValueAsText());
    assertEquals("1", root.get("request-id").getValueAsText());
    assertEquals(location, root.get("location").getValueAsText());
    // Trigger again
    response = trigger.handleRequest(params, Collections.<String, String>emptyMap(), jobManager);
    assertEquals(HttpResponseStatus.ACCEPTED, response.getStatus());
    assertEquals("application/json", response.headers().get(HttpHeaders.Names.CONTENT_TYPE));
    assertEquals(Integer.toString(response.content().readableBytes()), response.headers().get(HttpHeaders.Names.CONTENT_LENGTH));
    assertEquals(location, response.headers().get(HttpHeaders.Names.LOCATION));
    json = response.content().toString(Charset.forName("UTF-8"));
    root = new ObjectMapper().readTree(json);
    assertEquals("accepted", root.get("status").getValueAsText());
    assertEquals("1", root.get("request-id").getValueAsText());
    assertEquals(location, root.get("location").getValueAsText());
    // Only single actual request
    verify(jobManager).ask(eq(new CancelJobWithSavepoint(jobId, "custom-directory")), any(FiniteDuration.class));
    // Query progress
    params.put("requestId", "1");
    response = progress.handleRequest(params, Collections.<String, String>emptyMap(), jobManager);
    assertEquals(HttpResponseStatus.ACCEPTED, response.getStatus());
    assertEquals("application/json", response.headers().get(HttpHeaders.Names.CONTENT_TYPE));
    assertEquals(Integer.toString(response.content().readableBytes()), response.headers().get(HttpHeaders.Names.CONTENT_LENGTH));
    json = response.content().toString(Charset.forName("UTF-8"));
    root = new ObjectMapper().readTree(json);
    assertEquals("in-progress", root.get("status").getValueAsText());
    assertEquals("1", root.get("request-id").getValueAsText());
    // Complete
    promise.success(new CancellationSuccess(jobId, "_path-savepoint_"));
    response = progress.handleRequest(params, Collections.<String, String>emptyMap(), jobManager);
    assertEquals(HttpResponseStatus.CREATED, response.getStatus());
    assertEquals("application/json", response.headers().get(HttpHeaders.Names.CONTENT_TYPE));
    assertEquals(Integer.toString(response.content().readableBytes()), response.headers().get(HttpHeaders.Names.CONTENT_LENGTH));
    json = response.content().toString(Charset.forName("UTF-8"));
    root = new ObjectMapper().readTree(json);
    assertEquals("success", root.get("status").getValueAsText());
    assertEquals("1", root.get("request-id").getValueAsText());
    assertEquals("_path-savepoint_", root.get("savepoint-path").getValueAsText());
    // Query again, keep recent history
    response = progress.handleRequest(params, Collections.<String, String>emptyMap(), jobManager);
    assertEquals(HttpResponseStatus.CREATED, response.getStatus());
    assertEquals("application/json", response.headers().get(HttpHeaders.Names.CONTENT_TYPE));
    assertEquals(Integer.toString(response.content().readableBytes()), response.headers().get(HttpHeaders.Names.CONTENT_LENGTH));
    json = response.content().toString(Charset.forName("UTF-8"));
    root = new ObjectMapper().readTree(json);
    assertEquals("success", root.get("status").getValueAsText());
    assertEquals("1", root.get("request-id").getValueAsText());
    assertEquals("_path-savepoint_", root.get("savepoint-path").getValueAsText());
    // Query for unknown request
    params.put("requestId", "9929");
    response = progress.handleRequest(params, Collections.<String, String>emptyMap(), jobManager);
    assertEquals(HttpResponseStatus.BAD_REQUEST, response.getStatus());
    assertEquals("application/json", response.headers().get(HttpHeaders.Names.CONTENT_TYPE));
    assertEquals(Integer.toString(response.content().readableBytes()), response.headers().get(HttpHeaders.Names.CONTENT_LENGTH));
    json = response.content().toString(Charset.forName("UTF-8"));
    root = new ObjectMapper().readTree(json);
    assertEquals("failed", root.get("status").getValueAsText());
    assertEquals("9929", root.get("request-id").getValueAsText());
    assertEquals("Unknown job/request ID", root.get("cause").getValueAsText());
}

Example 8 with CancellationSuccess

use of org.apache.flink.runtime.messages.JobManagerMessages.CancellationSuccess in project flink by apache.

the class JobManagerTest method testSavepointRestoreSettings.

/**
	 * Tests that configured {@link SavepointRestoreSettings} are respected.
	 */
@Test
public void testSavepointRestoreSettings() throws Exception {
    FiniteDuration timeout = new FiniteDuration(30, TimeUnit.SECONDS);
    ActorSystem actorSystem = null;
    ActorGateway jobManager = null;
    ActorGateway archiver = null;
    ActorGateway taskManager = null;
    try {
        actorSystem = AkkaUtils.createLocalActorSystem(new Configuration());
        Tuple2<ActorRef, ActorRef> master = JobManager.startJobManagerActors(new Configuration(), actorSystem, TestingUtils.defaultExecutor(), TestingUtils.defaultExecutor(), Option.apply("jm"), Option.apply("arch"), TestingJobManager.class, TestingMemoryArchivist.class);
        jobManager = new AkkaActorGateway(master._1(), null);
        archiver = new AkkaActorGateway(master._2(), null);
        Configuration tmConfig = new Configuration();
        tmConfig.setInteger(ConfigConstants.TASK_MANAGER_NUM_TASK_SLOTS, 4);
        ActorRef taskManagerRef = TaskManager.startTaskManagerComponentsAndActor(tmConfig, ResourceID.generate(), actorSystem, "localhost", Option.apply("tm"), Option.<LeaderRetrievalService>apply(new StandaloneLeaderRetrievalService(jobManager.path())), true, TestingTaskManager.class);
        taskManager = new AkkaActorGateway(taskManagerRef, null);
        // Wait until connected
        Object msg = new TestingTaskManagerMessages.NotifyWhenRegisteredAtJobManager(jobManager.actor());
        Await.ready(taskManager.ask(msg, timeout), timeout);
        // Create job graph
        JobVertex sourceVertex = new JobVertex("Source");
        sourceVertex.setInvokableClass(BlockingStatefulInvokable.class);
        sourceVertex.setParallelism(1);
        JobGraph jobGraph = new JobGraph("TestingJob", sourceVertex);
        JobSnapshottingSettings snapshottingSettings = new JobSnapshottingSettings(Collections.singletonList(sourceVertex.getID()), Collections.singletonList(sourceVertex.getID()), Collections.singletonList(sourceVertex.getID()), // deactivated checkpointing
        Long.MAX_VALUE, 360000, 0, Integer.MAX_VALUE, ExternalizedCheckpointSettings.none(), null, true);
        jobGraph.setSnapshotSettings(snapshottingSettings);
        // Submit job graph
        msg = new JobManagerMessages.SubmitJob(jobGraph, ListeningBehaviour.DETACHED);
        Await.result(jobManager.ask(msg, timeout), timeout);
        // Wait for all tasks to be running
        msg = new TestingJobManagerMessages.WaitForAllVerticesToBeRunning(jobGraph.getJobID());
        Await.result(jobManager.ask(msg, timeout), timeout);
        // Trigger savepoint
        File targetDirectory = tmpFolder.newFolder();
        msg = new TriggerSavepoint(jobGraph.getJobID(), Option.apply(targetDirectory.getAbsolutePath()));
        Future<Object> future = jobManager.ask(msg, timeout);
        Object result = Await.result(future, timeout);
        String savepointPath = ((TriggerSavepointSuccess) result).savepointPath();
        // Cancel because of restarts
        msg = new TestingJobManagerMessages.NotifyWhenJobRemoved(jobGraph.getJobID());
        Future<?> removedFuture = jobManager.ask(msg, timeout);
        Future<?> cancelFuture = jobManager.ask(new CancelJob(jobGraph.getJobID()), timeout);
        Object response = Await.result(cancelFuture, timeout);
        assertTrue("Unexpected response: " + response, response instanceof CancellationSuccess);
        Await.ready(removedFuture, timeout);
        // Adjust the job (we need a new operator ID)
        JobVertex newSourceVertex = new JobVertex("NewSource");
        newSourceVertex.setInvokableClass(BlockingStatefulInvokable.class);
        newSourceVertex.setParallelism(1);
        JobGraph newJobGraph = new JobGraph("NewTestingJob", newSourceVertex);
        JobSnapshottingSettings newSnapshottingSettings = new JobSnapshottingSettings(Collections.singletonList(newSourceVertex.getID()), Collections.singletonList(newSourceVertex.getID()), Collections.singletonList(newSourceVertex.getID()), // deactivated checkpointing
        Long.MAX_VALUE, 360000, 0, Integer.MAX_VALUE, ExternalizedCheckpointSettings.none(), null, true);
        newJobGraph.setSnapshotSettings(newSnapshottingSettings);
        SavepointRestoreSettings restoreSettings = SavepointRestoreSettings.forPath(savepointPath, false);
        newJobGraph.setSavepointRestoreSettings(restoreSettings);
        msg = new JobManagerMessages.SubmitJob(newJobGraph, ListeningBehaviour.DETACHED);
        response = Await.result(jobManager.ask(msg, timeout), timeout);
        assertTrue("Unexpected response: " + response, response instanceof JobManagerMessages.JobResultFailure);
        JobManagerMessages.JobResultFailure failure = (JobManagerMessages.JobResultFailure) response;
        Throwable cause = failure.cause().deserializeError(ClassLoader.getSystemClassLoader());
        assertTrue(cause instanceof IllegalStateException);
        assertTrue(cause.getMessage().contains("allowNonRestoredState"));
        // Wait until removed
        msg = new TestingJobManagerMessages.NotifyWhenJobRemoved(newJobGraph.getJobID());
        Await.ready(jobManager.ask(msg, timeout), timeout);
        // Resubmit, but allow non restored state now
        restoreSettings = SavepointRestoreSettings.forPath(savepointPath, true);
        newJobGraph.setSavepointRestoreSettings(restoreSettings);
        msg = new JobManagerMessages.SubmitJob(newJobGraph, ListeningBehaviour.DETACHED);
        response = Await.result(jobManager.ask(msg, timeout), timeout);
        assertTrue("Unexpected response: " + response, response instanceof JobManagerMessages.JobSubmitSuccess);
    } finally {
        if (actorSystem != null) {
            actorSystem.shutdown();
        }
        if (archiver != null) {
            archiver.actor().tell(PoisonPill.getInstance(), ActorRef.noSender());
        }
        if (jobManager != null) {
            jobManager.actor().tell(PoisonPill.getInstance(), ActorRef.noSender());
        }
        if (taskManager != null) {
            taskManager.actor().tell(PoisonPill.getInstance(), ActorRef.noSender());
        }
    }
}

Example 9 with CancellationSuccess

use of org.apache.flink.runtime.messages.JobManagerMessages.CancellationSuccess in project flink by apache.

the class AbstractQueryableStateITCase method testQueryNonStartedJobState.

/**
	 * Similar tests as {@link #testValueState()} but before submitting the
	 * job, we already issue one request which fails.
	 */
@Test
public void testQueryNonStartedJobState() throws Exception {
    // Config
    final Deadline deadline = TEST_TIMEOUT.fromNow();
    final int numElements = 1024;
    final QueryableStateClient client = new QueryableStateClient(cluster.configuration());
    JobID jobId = null;
    try {
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.setStateBackend(stateBackend);
        env.setParallelism(NUM_SLOTS);
        // Very important, because cluster is shared between tests and we
        // don't explicitly check that all slots are available before
        // submitting.
        env.setRestartStrategy(RestartStrategies.fixedDelayRestart(Integer.MAX_VALUE, 1000));
        DataStream<Tuple2<Integer, Long>> source = env.addSource(new TestAscendingValueSource(numElements));
        // Value state
        ValueStateDescriptor<Tuple2<Integer, Long>> valueState = new ValueStateDescriptor<>("any", source.getType(), null);
        QueryableStateStream<Integer, Tuple2<Integer, Long>> queryableState = source.keyBy(new KeySelector<Tuple2<Integer, Long>, Integer>() {

            @Override
            public Integer getKey(Tuple2<Integer, Long> value) throws Exception {
                return value.f0;
            }
        }).asQueryableState("hakuna", valueState);
        // Submit the job graph
        JobGraph jobGraph = env.getStreamGraph().getJobGraph();
        jobId = jobGraph.getJobID();
        // Now query
        long expected = numElements;
        // query once
        client.getKvState(jobId, queryableState.getQueryableStateName(), 0, KvStateRequestSerializer.serializeKeyAndNamespace(0, queryableState.getKeySerializer(), VoidNamespace.INSTANCE, VoidNamespaceSerializer.INSTANCE));
        cluster.submitJobDetached(jobGraph);
        executeValueQuery(deadline, client, jobId, queryableState, expected);
    } finally {
        // Free cluster resources
        if (jobId != null) {
            Future<CancellationSuccess> cancellation = cluster.getLeaderGateway(deadline.timeLeft()).ask(new JobManagerMessages.CancelJob(jobId), deadline.timeLeft()).mapTo(ClassTag$.MODULE$.<CancellationSuccess>apply(CancellationSuccess.class));
            Await.ready(cancellation, deadline.timeLeft());
        }
        client.shutDown();
    }
}

Example 10 with CancellationSuccess

use of org.apache.flink.runtime.messages.JobManagerMessages.CancellationSuccess in project flink by apache.

the class AbstractQueryableStateITCase method testValueStateDefault.

/**
	 * Tests simple value state queryable state instance with a default value
	 * set. Each source emits (subtaskIndex, 0)..(subtaskIndex, numElements)
	 * tuples, the key is mapped to 1 but key 0 is queried which should throw
	 * a {@link UnknownKeyOrNamespace} exception.
	 *
	 * @throws UnknownKeyOrNamespace thrown due querying a non-existent key
	 */
@Test(expected = UnknownKeyOrNamespace.class)
public void testValueStateDefault() throws Exception, UnknownKeyOrNamespace {
    // Config
    final Deadline deadline = TEST_TIMEOUT.fromNow();
    final int numElements = 1024;
    final QueryableStateClient client = new QueryableStateClient(cluster.configuration());
    JobID jobId = null;
    try {
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.setStateBackend(stateBackend);
        env.setParallelism(NUM_SLOTS);
        // Very important, because cluster is shared between tests and we
        // don't explicitly check that all slots are available before
        // submitting.
        env.setRestartStrategy(RestartStrategies.fixedDelayRestart(Integer.MAX_VALUE, 1000));
        DataStream<Tuple2<Integer, Long>> source = env.addSource(new TestAscendingValueSource(numElements));
        // Value state
        ValueStateDescriptor<Tuple2<Integer, Long>> valueState = new ValueStateDescriptor<>("any", source.getType(), Tuple2.of(0, 1337l));
        // only expose key "1"
        QueryableStateStream<Integer, Tuple2<Integer, Long>> queryableState = source.keyBy(new KeySelector<Tuple2<Integer, Long>, Integer>() {

            @Override
            public Integer getKey(Tuple2<Integer, Long> value) throws Exception {
                return 1;
            }
        }).asQueryableState("hakuna", valueState);
        // Submit the job graph
        JobGraph jobGraph = env.getStreamGraph().getJobGraph();
        jobId = jobGraph.getJobID();
        cluster.submitJobDetached(jobGraph);
        // Now query
        int key = 0;
        final byte[] serializedKey = KvStateRequestSerializer.serializeKeyAndNamespace(key, queryableState.getKeySerializer(), VoidNamespace.INSTANCE, VoidNamespaceSerializer.INSTANCE);
        Future<byte[]> future = getKvStateWithRetries(client, jobId, queryableState.getQueryableStateName(), key, serializedKey, QUERY_RETRY_DELAY, true);
        Await.result(future, deadline.timeLeft());
    } finally {
        // Free cluster resources
        if (jobId != null) {
            Future<CancellationSuccess> cancellation = cluster.getLeaderGateway(deadline.timeLeft()).ask(new JobManagerMessages.CancelJob(jobId), deadline.timeLeft()).mapTo(ClassTag$.MODULE$.<CancellationSuccess>apply(CancellationSuccess.class));
            Await.ready(cancellation, deadline.timeLeft());
        }
        client.shutDown();
    }
}