Search in sources :

Example 1 with ProgressRequestCallback

use of org.apache.beam.fn.harness.PTransformRunnerFactory.ProgressRequestCallback in project beam by apache.

the class ProcessBundleHandler method monitoringData.

private ImmutableMap<String, ByteString> monitoringData(BundleProcessor bundleProcessor) throws Exception {
    ImmutableMap.Builder<String, ByteString> result = ImmutableMap.builder();
    // Get start bundle Execution Time Metrics.
    result.putAll(bundleProcessor.getStartFunctionRegistry().getExecutionTimeMonitoringData(shortIds));
    // Get process bundle Execution Time Metrics.
    result.putAll(bundleProcessor.getpCollectionConsumerRegistry().getExecutionTimeMonitoringData(shortIds));
    // Get finish bundle Execution Time Metrics.
    result.putAll(bundleProcessor.getFinishFunctionRegistry().getExecutionTimeMonitoringData(shortIds));
    // Extract MonitoringInfos that come from the metrics container registry.
    result.putAll(bundleProcessor.getMetricsContainerRegistry().getMonitoringData(shortIds));
    // Add any additional monitoring infos that the "runners" report explicitly.
    for (ProgressRequestCallback progressRequestCallback : bundleProcessor.getProgressRequestCallbacks()) {
        // TODO(BEAM-6597): Plumb reporting monitoring infos using the short id system upstream.
        for (MetricsApi.MonitoringInfo monitoringInfo : progressRequestCallback.getMonitoringInfos()) {
            ByteString payload = monitoringInfo.getPayload();
            String shortId = shortIds.getOrCreateShortId(monitoringInfo.toBuilder().clearPayload().build());
            result.put(shortId, payload);
        }
    }
    return result.build();
}
Also used : ByteString(org.apache.beam.vendor.grpc.v1p43p2.com.google.protobuf.ByteString) ProgressRequestCallback(org.apache.beam.fn.harness.PTransformRunnerFactory.ProgressRequestCallback) ByteString(org.apache.beam.vendor.grpc.v1p43p2.com.google.protobuf.ByteString) MetricsApi(org.apache.beam.model.pipeline.v1.MetricsApi) ImmutableMap(org.apache.beam.vendor.guava.v26_0_jre.com.google.common.collect.ImmutableMap)

Example 2 with ProgressRequestCallback

use of org.apache.beam.fn.harness.PTransformRunnerFactory.ProgressRequestCallback in project beam by apache.

the class ProcessBundleHandler method createBundleProcessor.

private BundleProcessor createBundleProcessor(String bundleId, BeamFnApi.ProcessBundleRequest processBundleRequest) throws IOException {
    BeamFnApi.ProcessBundleDescriptor bundleDescriptor = fnApiRegistry.apply(bundleId);
    SetMultimap<String, String> pCollectionIdsToConsumingPTransforms = HashMultimap.create();
    MetricsContainerStepMap metricsContainerRegistry = new MetricsContainerStepMap();
    ExecutionStateTracker stateTracker = new ExecutionStateTracker(ExecutionStateSampler.instance());
    PCollectionConsumerRegistry pCollectionConsumerRegistry = new PCollectionConsumerRegistry(metricsContainerRegistry, stateTracker);
    HashSet<String> processedPTransformIds = new HashSet<>();
    PTransformFunctionRegistry startFunctionRegistry = new PTransformFunctionRegistry(metricsContainerRegistry, stateTracker, ExecutionStateTracker.START_STATE_NAME);
    PTransformFunctionRegistry finishFunctionRegistry = new PTransformFunctionRegistry(metricsContainerRegistry, stateTracker, ExecutionStateTracker.FINISH_STATE_NAME);
    List<ThrowingRunnable> resetFunctions = new ArrayList<>();
    List<ThrowingRunnable> tearDownFunctions = new ArrayList<>();
    List<ProgressRequestCallback> progressRequestCallbacks = new ArrayList<>();
    // Build a multimap of PCollection ids to PTransform ids which consume said PCollections
    for (Map.Entry<String, RunnerApi.PTransform> entry : bundleDescriptor.getTransformsMap().entrySet()) {
        for (String pCollectionId : entry.getValue().getInputsMap().values()) {
            pCollectionIdsToConsumingPTransforms.put(pCollectionId, entry.getKey());
        }
    }
    // Instantiate a State API call handler depending on whether a State ApiServiceDescriptor was
    // specified.
    HandleStateCallsForBundle beamFnStateClient;
    if (bundleDescriptor.hasStateApiServiceDescriptor()) {
        BeamFnStateClient underlyingClient = beamFnStateGrpcClientCache.forApiServiceDescriptor(bundleDescriptor.getStateApiServiceDescriptor());
        beamFnStateClient = new BlockTillStateCallsFinish(underlyingClient);
    } else {
        beamFnStateClient = new FailAllStateCallsForBundle(processBundleRequest);
    }
    BundleSplitListener.InMemory splitListener = BundleSplitListener.InMemory.create();
    Collection<CallbackRegistration> bundleFinalizationCallbackRegistrations = new ArrayList<>();
    BundleFinalizer bundleFinalizer = new BundleFinalizer() {

        @Override
        public void afterBundleCommit(Instant callbackExpiry, Callback callback) {
            bundleFinalizationCallbackRegistrations.add(CallbackRegistration.create(callbackExpiry, callback));
        }
    };
    BundleProcessor bundleProcessor = BundleProcessor.create(processWideCache, bundleDescriptor, startFunctionRegistry, finishFunctionRegistry, resetFunctions, tearDownFunctions, progressRequestCallbacks, splitListener, pCollectionConsumerRegistry, metricsContainerRegistry, stateTracker, beamFnStateClient, bundleFinalizationCallbackRegistrations, runnerCapabilities);
    // Create a BeamFnStateClient
    for (Map.Entry<String, RunnerApi.PTransform> entry : bundleDescriptor.getTransformsMap().entrySet()) {
        // TODO: Remove source as a root and have it be triggered by the Runner.
        if (!DATA_INPUT_URN.equals(entry.getValue().getSpec().getUrn()) && !DATA_OUTPUT_URN.equals(entry.getValue().getSpec().getUrn()) && !JAVA_SOURCE_URN.equals(entry.getValue().getSpec().getUrn()) && !PTransformTranslation.READ_TRANSFORM_URN.equals(entry.getValue().getSpec().getUrn())) {
            continue;
        }
        createRunnerAndConsumersForPTransformRecursively(beamFnStateClient, beamFnDataClient, entry.getKey(), entry.getValue(), bundleProcessor::getInstructionId, bundleProcessor::getCacheTokens, bundleProcessor::getBundleCache, bundleDescriptor, pCollectionIdsToConsumingPTransforms, pCollectionConsumerRegistry, processedPTransformIds, startFunctionRegistry, finishFunctionRegistry, resetFunctions::add, tearDownFunctions::add, (apiServiceDescriptor, dataEndpoint) -> {
            if (!bundleProcessor.getInboundEndpointApiServiceDescriptors().contains(apiServiceDescriptor)) {
                bundleProcessor.getInboundEndpointApiServiceDescriptors().add(apiServiceDescriptor);
            }
            bundleProcessor.getInboundDataEndpoints().add(dataEndpoint);
        }, (timerEndpoint) -> {
            if (!bundleDescriptor.hasTimerApiServiceDescriptor()) {
                throw new IllegalStateException(String.format("Timers are unsupported because the " + "ProcessBundleRequest %s does not provide a timer ApiServiceDescriptor.", bundleId));
            }
            bundleProcessor.getTimerEndpoints().add(timerEndpoint);
        }, progressRequestCallbacks::add, splitListener, bundleFinalizer, bundleProcessor.getChannelRoots(), bundleProcessor.getOutboundAggregators(), bundleProcessor.getRunnerCapabilities());
    }
    bundleProcessor.finish();
    return bundleProcessor;
}
Also used : MetricsContainerStepMap(org.apache.beam.runners.core.metrics.MetricsContainerStepMap) BeamFnStateClient(org.apache.beam.fn.harness.state.BeamFnStateClient) ProgressRequestCallback(org.apache.beam.fn.harness.PTransformRunnerFactory.ProgressRequestCallback) ArrayList(java.util.ArrayList) ByteString(org.apache.beam.vendor.grpc.v1p43p2.com.google.protobuf.ByteString) PCollectionConsumerRegistry(org.apache.beam.fn.harness.data.PCollectionConsumerRegistry) ThrowingRunnable(org.apache.beam.sdk.function.ThrowingRunnable) CallbackRegistration(org.apache.beam.fn.harness.control.FinalizeBundleHandler.CallbackRegistration) ExecutionStateTracker(org.apache.beam.runners.core.metrics.ExecutionStateTracker) HashSet(java.util.HashSet) PTransform(org.apache.beam.model.pipeline.v1.RunnerApi.PTransform) ProcessBundleDescriptor(org.apache.beam.model.fnexecution.v1.BeamFnApi.ProcessBundleDescriptor) BeamFnApi(org.apache.beam.model.fnexecution.v1.BeamFnApi) PTransformFunctionRegistry(org.apache.beam.fn.harness.data.PTransformFunctionRegistry) Instant(org.joda.time.Instant) BundleFinalizer(org.apache.beam.sdk.transforms.DoFn.BundleFinalizer) ProgressRequestCallback(org.apache.beam.fn.harness.PTransformRunnerFactory.ProgressRequestCallback) ImmutableMap(org.apache.beam.vendor.guava.v26_0_jre.com.google.common.collect.ImmutableMap) MetricsContainerStepMap(org.apache.beam.runners.core.metrics.MetricsContainerStepMap) Map(java.util.Map) ShortIdMap(org.apache.beam.runners.core.metrics.ShortIdMap) LinkedHashMap(java.util.LinkedHashMap) WeakHashMap(java.util.WeakHashMap)

Aggregations

ProgressRequestCallback (org.apache.beam.fn.harness.PTransformRunnerFactory.ProgressRequestCallback)2 ByteString (org.apache.beam.vendor.grpc.v1p43p2.com.google.protobuf.ByteString)2 ImmutableMap (org.apache.beam.vendor.guava.v26_0_jre.com.google.common.collect.ImmutableMap)2 ArrayList (java.util.ArrayList)1 HashSet (java.util.HashSet)1 LinkedHashMap (java.util.LinkedHashMap)1 Map (java.util.Map)1 WeakHashMap (java.util.WeakHashMap)1 CallbackRegistration (org.apache.beam.fn.harness.control.FinalizeBundleHandler.CallbackRegistration)1 PCollectionConsumerRegistry (org.apache.beam.fn.harness.data.PCollectionConsumerRegistry)1 PTransformFunctionRegistry (org.apache.beam.fn.harness.data.PTransformFunctionRegistry)1 BeamFnStateClient (org.apache.beam.fn.harness.state.BeamFnStateClient)1 BeamFnApi (org.apache.beam.model.fnexecution.v1.BeamFnApi)1 ProcessBundleDescriptor (org.apache.beam.model.fnexecution.v1.BeamFnApi.ProcessBundleDescriptor)1 MetricsApi (org.apache.beam.model.pipeline.v1.MetricsApi)1 PTransform (org.apache.beam.model.pipeline.v1.RunnerApi.PTransform)1 ExecutionStateTracker (org.apache.beam.runners.core.metrics.ExecutionStateTracker)1 MetricsContainerStepMap (org.apache.beam.runners.core.metrics.MetricsContainerStepMap)1 ShortIdMap (org.apache.beam.runners.core.metrics.ShortIdMap)1 ThrowingRunnable (org.apache.beam.sdk.function.ThrowingRunnable)1