Search in sources :

Example 1 with AlertPublisherContext

use of io.cdap.cdap.etl.api.AlertPublisherContext 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();
}
Also used : PluginFunctionContext(io.cdap.cdap.etl.spark.function.PluginFunctionContext) AlertPublisher(io.cdap.cdap.etl.api.AlertPublisher) PipelineRuntime(io.cdap.cdap.etl.common.PipelineRuntime) SparkPipelineRuntime(io.cdap.cdap.etl.spark.SparkPipelineRuntime) SparkPipelineRuntime(io.cdap.cdap.etl.spark.SparkPipelineRuntime) TrackedIterator(io.cdap.cdap.etl.common.TrackedIterator) Alert(io.cdap.cdap.etl.api.Alert) AlertPublisherContext(io.cdap.cdap.etl.api.AlertPublisherContext) DefaultAlertPublisherContext(io.cdap.cdap.etl.common.DefaultAlertPublisherContext) DefaultAlertPublisherContext(io.cdap.cdap.etl.common.DefaultAlertPublisherContext) StageMetrics(io.cdap.cdap.etl.api.StageMetrics) DefaultStageMetrics(io.cdap.cdap.etl.common.DefaultStageMetrics) DefaultStageMetrics(io.cdap.cdap.etl.common.DefaultStageMetrics) JavaRDD(org.apache.spark.api.java.JavaRDD)

Example 2 with AlertPublisherContext

use of io.cdap.cdap.etl.api.AlertPublisherContext 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);
    }
}
Also used : NodeValue(io.cdap.cdap.api.workflow.NodeValue) PipelineRuntime(io.cdap.cdap.etl.common.PipelineRuntime) DefaultMacroEvaluator(io.cdap.cdap.etl.common.DefaultMacroEvaluator) MacroEvaluator(io.cdap.cdap.api.macro.MacroEvaluator) BatchActionContext(io.cdap.cdap.etl.api.batch.BatchActionContext) WorkflowBackedActionContext(io.cdap.cdap.etl.batch.WorkflowBackedActionContext) HashMap(java.util.HashMap) LinkedHashMap(java.util.LinkedHashMap) WorkflowToken(io.cdap.cdap.api.workflow.WorkflowToken) FieldOperation(io.cdap.cdap.etl.api.lineage.field.FieldOperation) Operation(io.cdap.cdap.api.lineage.field.Operation) AlertReader(io.cdap.cdap.etl.batch.connector.AlertReader) StageSpec(io.cdap.cdap.etl.proto.v2.spec.StageSpec) DefaultMacroEvaluator(io.cdap.cdap.etl.common.DefaultMacroEvaluator) List(java.util.List) ArrayList(java.util.ArrayList) LinkedList(java.util.LinkedList) StageMetrics(io.cdap.cdap.etl.api.StageMetrics) DefaultStageMetrics(io.cdap.cdap.etl.common.DefaultStageMetrics) AlertPublisherContext(io.cdap.cdap.etl.api.AlertPublisherContext) DefaultAlertPublisherContext(io.cdap.cdap.etl.common.DefaultAlertPublisherContext) FieldLineageProcessor(io.cdap.cdap.etl.lineage.FieldLineageProcessor) AlertPublisher(io.cdap.cdap.etl.api.AlertPublisher) FileSet(io.cdap.cdap.api.dataset.lib.FileSet) TrackedIterator(io.cdap.cdap.etl.common.TrackedIterator) WorkflowContext(io.cdap.cdap.api.workflow.WorkflowContext) DisjointConnectionsException(io.cdap.cdap.etl.planner.DisjointConnectionsException) ValidationException(io.cdap.cdap.etl.api.validation.ValidationException) Alert(io.cdap.cdap.etl.api.Alert) PostAction(io.cdap.cdap.etl.api.batch.PostAction) DefaultAlertPublisherContext(io.cdap.cdap.etl.common.DefaultAlertPublisherContext) Map(java.util.Map) HashMap(java.util.HashMap) LinkedHashMap(java.util.LinkedHashMap) DefaultStageMetrics(io.cdap.cdap.etl.common.DefaultStageMetrics) ProgramStatus(io.cdap.cdap.api.ProgramStatus)

Example 3 with AlertPublisherContext

use of io.cdap.cdap.etl.api.AlertPublisherContext in project cdap by caskdata.

the class StreamingAlertPublishFunction method call.

@Override
public void call(JavaRDD<Alert> data, Time batchTime) throws Exception {
    MacroEvaluator evaluator = new DefaultMacroEvaluator(new BasicArguments(sec), batchTime.milliseconds(), sec.getSecureStore(), sec.getServiceDiscoverer(), sec.getNamespace());
    PluginContext pluginContext = new SparkPipelinePluginContext(sec.getPluginContext(), sec.getMetrics(), stageSpec.isStageLoggingEnabled(), stageSpec.isProcessTimingEnabled());
    String stageName = stageSpec.getName();
    AlertPublisher alertPublisher = pluginContext.newPluginInstance(stageName, evaluator);
    PipelineRuntime pipelineRuntime = new SparkPipelineRuntime(sec, batchTime.milliseconds());
    AlertPublisherContext alertPublisherContext = new DefaultAlertPublisherContext(pipelineRuntime, stageSpec, sec.getMessagingContext(), sec.getAdmin());
    alertPublisher.initialize(alertPublisherContext);
    StageMetrics stageMetrics = new DefaultStageMetrics(sec.getMetrics(), stageName);
    TrackedIterator<Alert> trackedAlerts = new TrackedIterator<>(data.collect().iterator(), stageMetrics, Constants.Metrics.RECORDS_IN);
    alertPublisher.publish(trackedAlerts);
    alertPublisher.destroy();
}
Also used : DefaultMacroEvaluator(io.cdap.cdap.etl.common.DefaultMacroEvaluator) MacroEvaluator(io.cdap.cdap.api.macro.MacroEvaluator) AlertPublisher(io.cdap.cdap.etl.api.AlertPublisher) SparkPipelineRuntime(io.cdap.cdap.etl.spark.SparkPipelineRuntime) PipelineRuntime(io.cdap.cdap.etl.common.PipelineRuntime) SparkPipelinePluginContext(io.cdap.cdap.etl.spark.plugin.SparkPipelinePluginContext) PluginContext(io.cdap.cdap.api.plugin.PluginContext) SparkPipelineRuntime(io.cdap.cdap.etl.spark.SparkPipelineRuntime) TrackedIterator(io.cdap.cdap.etl.common.TrackedIterator) SparkPipelinePluginContext(io.cdap.cdap.etl.spark.plugin.SparkPipelinePluginContext) DefaultMacroEvaluator(io.cdap.cdap.etl.common.DefaultMacroEvaluator) Alert(io.cdap.cdap.etl.api.Alert) BasicArguments(io.cdap.cdap.etl.common.BasicArguments) AlertPublisherContext(io.cdap.cdap.etl.api.AlertPublisherContext) DefaultAlertPublisherContext(io.cdap.cdap.etl.common.DefaultAlertPublisherContext) DefaultAlertPublisherContext(io.cdap.cdap.etl.common.DefaultAlertPublisherContext) DefaultStageMetrics(io.cdap.cdap.etl.common.DefaultStageMetrics) StageMetrics(io.cdap.cdap.etl.api.StageMetrics) DefaultStageMetrics(io.cdap.cdap.etl.common.DefaultStageMetrics)

Aggregations

Alert (io.cdap.cdap.etl.api.Alert)3 AlertPublisher (io.cdap.cdap.etl.api.AlertPublisher)3 AlertPublisherContext (io.cdap.cdap.etl.api.AlertPublisherContext)3 StageMetrics (io.cdap.cdap.etl.api.StageMetrics)3 DefaultAlertPublisherContext (io.cdap.cdap.etl.common.DefaultAlertPublisherContext)3 DefaultStageMetrics (io.cdap.cdap.etl.common.DefaultStageMetrics)3 PipelineRuntime (io.cdap.cdap.etl.common.PipelineRuntime)3 TrackedIterator (io.cdap.cdap.etl.common.TrackedIterator)3 MacroEvaluator (io.cdap.cdap.api.macro.MacroEvaluator)2 DefaultMacroEvaluator (io.cdap.cdap.etl.common.DefaultMacroEvaluator)2 SparkPipelineRuntime (io.cdap.cdap.etl.spark.SparkPipelineRuntime)2 ProgramStatus (io.cdap.cdap.api.ProgramStatus)1 FileSet (io.cdap.cdap.api.dataset.lib.FileSet)1 Operation (io.cdap.cdap.api.lineage.field.Operation)1 PluginContext (io.cdap.cdap.api.plugin.PluginContext)1 NodeValue (io.cdap.cdap.api.workflow.NodeValue)1 WorkflowContext (io.cdap.cdap.api.workflow.WorkflowContext)1 WorkflowToken (io.cdap.cdap.api.workflow.WorkflowToken)1 BatchActionContext (io.cdap.cdap.etl.api.batch.BatchActionContext)1 PostAction (io.cdap.cdap.etl.api.batch.PostAction)1