Examples with PipelineRuntime io.cdap.cdap.etl.common.PipelineRuntime used on opensource projects

Example 1 with PipelineRuntime

use of io.cdap.cdap.etl.common.PipelineRuntime in project cdap by caskdata.

the class BaseRDDCollection method createStoreTask.

@Override
public Runnable createStoreTask(final StageSpec stageSpec, final SparkSink<T> sink) throws Exception {
    return new Runnable() {

        @Override
        public void run() {
            String stageName = stageSpec.getName();
            PipelineRuntime pipelineRuntime = new SparkPipelineRuntime(sec);
            SparkExecutionPluginContext sparkPluginContext = new BasicSparkExecutionPluginContext(sec, jsc, datasetContext, pipelineRuntime, stageSpec);
            JavaRDD<T> countedRDD = rdd.map(new CountingFunction<T>(stageName, sec.getMetrics(), Constants.Metrics.RECORDS_IN, null));
            SparkConf sparkConf = jsc.getConf();
            try {
                sink.run(sparkPluginContext, countedRDD);
            } catch (Exception e) {
                throw Throwables.propagate(e);
            }
        }
    };
}

Example 2 with PipelineRuntime

use of io.cdap.cdap.etl.common.PipelineRuntime in project cdap by caskdata.

the class BaseRDDCollection method publishAlerts.

@Override
public void publishAlerts(StageSpec stageSpec, StageStatisticsCollector collector) throws Exception {
    PluginFunctionContext pluginFunctionContext = new PluginFunctionContext(stageSpec, sec, collector);
    AlertPublisher alertPublisher = pluginFunctionContext.createPlugin();
    PipelineRuntime pipelineRuntime = new SparkPipelineRuntime(sec);
    AlertPublisherContext alertPublisherContext = new DefaultAlertPublisherContext(pipelineRuntime, stageSpec, sec.getMessagingContext(), sec.getAdmin());
    alertPublisher.initialize(alertPublisherContext);
    StageMetrics stageMetrics = new DefaultStageMetrics(sec.getMetrics(), stageSpec.getName());
    TrackedIterator<Alert> trackedAlerts = new TrackedIterator<>(((JavaRDD<Alert>) rdd).collect().iterator(), stageMetrics, Constants.Metrics.RECORDS_IN);
    alertPublisher.publish(trackedAlerts);
    alertPublisher.destroy();
}

Example 3 with PipelineRuntime

use of io.cdap.cdap.etl.common.PipelineRuntime in project cdap by caskdata.

the class BaseRDDCollection method compute.

@Override
public <U> SparkCollection<U> compute(StageSpec stageSpec, SparkCompute<T, U> compute) throws Exception {
    String stageName = stageSpec.getName();
    PipelineRuntime pipelineRuntime = new SparkPipelineRuntime(sec);
    SparkExecutionPluginContext sparkPluginContext = new BasicSparkExecutionPluginContext(sec, jsc, datasetContext, pipelineRuntime, stageSpec);
    compute.initialize(sparkPluginContext);
    JavaRDD<T> countedInput = rdd.map(new CountingFunction<T>(stageName, sec.getMetrics(), Constants.Metrics.RECORDS_IN, null));
    SparkConf sparkConf = jsc.getConf();
    return wrap(compute.transform(sparkPluginContext, countedInput).map(new CountingFunction<U>(stageName, sec.getMetrics(), Constants.Metrics.RECORDS_OUT, sec.getDataTracer(stageName))));
}

Example 4 with PipelineRuntime

use of io.cdap.cdap.etl.common.PipelineRuntime in project cdap by caskdata.

the class SmartWorkflow method destroy.

@Override
public void destroy() {
    WorkflowContext workflowContext = getContext();
    PipelineRuntime pipelineRuntime = new PipelineRuntime(workflowContext, workflowMetrics);
    // Execute the post actions only if pipeline is not running in preview mode.
    if (!workflowContext.getDataTracer(PostAction.PLUGIN_TYPE).isEnabled()) {
        for (Map.Entry<String, PostAction> endingActionEntry : postActions.entrySet()) {
            String name = endingActionEntry.getKey();
            PostAction action = endingActionEntry.getValue();
            StageSpec stageSpec = stageSpecs.get(name);
            BatchActionContext context = new WorkflowBackedActionContext(workflowContext, pipelineRuntime, stageSpec);
            try {
                action.run(context);
            } catch (Throwable t) {
                LOG.error("Error while running post action {}.", name, t);
            }
        }
    }
    Map<String, String> connectorDatasets = GSON.fromJson(workflowContext.getWorkflowSpecification().getProperty(Constants.CONNECTOR_DATASETS), STAGE_DATASET_MAP);
    // publish all alerts
    for (Map.Entry<String, AlertPublisher> alertPublisherEntry : alertPublishers.entrySet()) {
        String stageName = alertPublisherEntry.getKey();
        AlertPublisher alertPublisher = alertPublisherEntry.getValue();
        FileSet alertConnector = workflowContext.getDataset(connectorDatasets.get(stageName));
        try (CloseableIterator<Alert> alerts = new AlertReader(alertConnector)) {
            if (!alerts.hasNext()) {
                continue;
            }
            StageMetrics stageMetrics = new DefaultStageMetrics(workflowMetrics, stageName);
            StageSpec stageSpec = stageSpecs.get(stageName);
            AlertPublisherContext alertContext = new DefaultAlertPublisherContext(pipelineRuntime, stageSpec, workflowContext, workflowContext.getAdmin());
            alertPublisher.initialize(alertContext);
            TrackedIterator<Alert> trackedIterator = new TrackedIterator<>(alerts, stageMetrics, Constants.Metrics.RECORDS_IN);
            alertPublisher.publish(trackedIterator);
        } catch (Exception e) {
            LOG.warn("Stage {} had errors publishing alerts. Alerts may not have been published.", stageName, e);
        } finally {
            try {
                alertPublisher.destroy();
            } catch (Exception e) {
                LOG.warn("Error destroying alert publisher for stage {}", stageName, e);
            }
        }
    }
    ProgramStatus status = getContext().getState().getStatus();
    if (status == ProgramStatus.FAILED) {
        WRAPPERLOGGER.error("Pipeline '{}' failed.", getContext().getApplicationSpecification().getName());
    } else {
        WRAPPERLOGGER.info("Pipeline '{}' {}.", getContext().getApplicationSpecification().getName(), status == ProgramStatus.COMPLETED ? "succeeded" : status.name().toLowerCase());
    }
    MacroEvaluator macroEvaluator = new DefaultMacroEvaluator(pipelineRuntime.getArguments(), workflowContext.getLogicalStartTime(), workflowContext, workflowContext, workflowContext.getNamespace());
    // Get resolved plugin properties
    Map<String, Map<String, String>> resolvedProperties = new HashMap<>();
    for (StageSpec spec : stageSpecs.values()) {
        String stageName = spec.getName();
        resolvedProperties.put(stageName, workflowContext.getPluginProperties(stageName, macroEvaluator).getProperties());
    }
    // Add resolved plugin properties to workflow token as a JSON String
    workflowContext.getToken().put(RESOLVED_PLUGIN_PROPERTIES_MAP, GSON.toJson(resolvedProperties));
    // record only if the Workflow is successful
    if (status != ProgramStatus.COMPLETED) {
        return;
    }
    // Collect field operations from each phase
    WorkflowToken token = workflowContext.getToken();
    List<NodeValue> allNodeValues = token.getAll(Constants.FIELD_OPERATION_KEY_IN_WORKFLOW_TOKEN);
    if (allNodeValues.isEmpty()) {
        // no field lineage recorded by any stage
        return;
    }
    Map<String, List<FieldOperation>> allStageOperations = new HashMap<>();
    for (StageSpec stageSpec : stageSpecs.values()) {
        allStageOperations.put(stageSpec.getName(), new ArrayList<>());
    }
    for (NodeValue nodeValue : allNodeValues) {
        Map<String, List<FieldOperation>> stageOperations = GSON.fromJson(nodeValue.getValue().toString(), STAGE_OPERATIONS_MAP);
        for (Map.Entry<String, List<FieldOperation>> entry : stageOperations.entrySet()) {
            // ignore them
            if (allStageOperations.containsKey(entry.getKey())) {
                allStageOperations.get(entry.getKey()).addAll(entry.getValue());
            }
        }
    }
    FieldLineageProcessor processor = new FieldLineageProcessor(spec);
    Set<Operation> processedOperations = processor.validateAndConvert(allStageOperations);
    if (!processedOperations.isEmpty()) {
        workflowContext.record(processedOperations);
    }
}

Example 5 with PipelineRuntime

use of io.cdap.cdap.etl.common.PipelineRuntime in project cdap by caskdata.

the class ETLSpark method initialize.

@Override
@TransactionPolicy(TransactionControl.EXPLICIT)
public void initialize() throws Exception {
    SparkClientContext context = getContext();
    SparkConf sparkConf = new SparkConf();
    sparkConf.set("spark.speculation", "false");
    // turn off auto-broadcast by default until we better understand the implications and can set this to a
    // value that we are confident is safe.
    sparkConf.set("spark.sql.autoBroadcastJoinThreshold", "-1");
    sparkConf.set("spark.maxRemoteBlockSizeFetchToMem", String.valueOf(Integer.MAX_VALUE - 512));
    sparkConf.set("spark.network.timeout", "600s");
    // Disable yarn app retries since spark already performs retries at a task level.
    sparkConf.set("spark.yarn.maxAppAttempts", "1");
    // to make sure fields that are the same but different casing are treated as different fields in auto-joins
    // see CDAP-17024
    sparkConf.set("spark.sql.caseSensitive", "true");
    context.setSparkConf(sparkConf);
    Map<String, String> properties = context.getSpecification().getProperties();
    BatchPhaseSpec phaseSpec = GSON.fromJson(properties.get(Constants.PIPELINEID), BatchPhaseSpec.class);
    for (Map.Entry<String, String> pipelineProperty : phaseSpec.getPipelineProperties().entrySet()) {
        sparkConf.set(pipelineProperty.getKey(), pipelineProperty.getValue());
    }
    PipelineRuntime pipelineRuntime = new PipelineRuntime(context);
    MacroEvaluator evaluator = new DefaultMacroEvaluator(pipelineRuntime.getArguments(), context.getLogicalStartTime(), context, context, context.getNamespace());
    SparkPreparer preparer = new SparkPreparer(context, context.getMetrics(), evaluator, pipelineRuntime);
    List<Finisher> finishers = preparer.prepare(phaseSpec);
    finisher = new CompositeFinisher(finishers);
}