Search in sources :

Example 1 with FeedIntakeOperatorDescriptor

use of org.apache.asterix.external.operators.FeedIntakeOperatorDescriptor in project asterixdb by apache.

the class FeedOperations method combineIntakeCollectJobs.

private static JobSpecification combineIntakeCollectJobs(MetadataProvider metadataProvider, Feed feed, JobSpecification intakeJob, List<JobSpecification> jobsList, List<FeedConnection> feedConnections, String[] intakeLocations) throws AlgebricksException, HyracksDataException {
    JobSpecification jobSpec = new JobSpecification(intakeJob.getFrameSize());
    // copy ingestor
    FeedIntakeOperatorDescriptor firstOp = (FeedIntakeOperatorDescriptor) intakeJob.getOperatorMap().get(new OperatorDescriptorId(0));
    FeedIntakeOperatorDescriptor ingestionOp;
    if (firstOp.getAdaptorFactory() == null) {
        ingestionOp = new FeedIntakeOperatorDescriptor(jobSpec, feed, firstOp.getAdaptorLibraryName(), firstOp.getAdaptorFactoryClassName(), firstOp.getAdapterOutputType(), firstOp.getPolicyAccessor(), firstOp.getOutputRecordDescriptors()[0]);
    } else {
        ingestionOp = new FeedIntakeOperatorDescriptor(jobSpec, feed, firstOp.getAdaptorFactory(), firstOp.getAdapterOutputType(), firstOp.getPolicyAccessor(), firstOp.getOutputRecordDescriptors()[0]);
    }
    // create replicator
    ReplicateOperatorDescriptor replicateOp = new ReplicateOperatorDescriptor(jobSpec, ingestionOp.getOutputRecordDescriptors()[0], jobsList.size());
    jobSpec.connect(new OneToOneConnectorDescriptor(jobSpec), ingestionOp, 0, replicateOp, 0);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(jobSpec, ingestionOp, intakeLocations);
    PartitionConstraintHelper.addAbsoluteLocationConstraint(jobSpec, replicateOp, intakeLocations);
    // Loop over the jobs to copy operators and connections
    Map<OperatorDescriptorId, OperatorDescriptorId> operatorIdMapping = new HashMap<>();
    Map<ConnectorDescriptorId, ConnectorDescriptorId> connectorIdMapping = new HashMap<>();
    Map<OperatorDescriptorId, List<LocationConstraint>> operatorLocations = new HashMap<>();
    Map<OperatorDescriptorId, Integer> operatorCounts = new HashMap<>();
    List<JobId> jobIds = new ArrayList<>();
    FeedMetaOperatorDescriptor metaOp;
    for (int iter1 = 0; iter1 < jobsList.size(); iter1++) {
        FeedConnection curFeedConnection = feedConnections.get(iter1);
        JobSpecification subJob = jobsList.get(iter1);
        operatorIdMapping.clear();
        Map<OperatorDescriptorId, IOperatorDescriptor> operatorsMap = subJob.getOperatorMap();
        String datasetName = feedConnections.get(iter1).getDatasetName();
        FeedConnectionId feedConnectionId = new FeedConnectionId(ingestionOp.getEntityId(), datasetName);
        FeedPolicyEntity feedPolicyEntity = FeedMetadataUtil.validateIfPolicyExists(curFeedConnection.getDataverseName(), curFeedConnection.getPolicyName(), metadataProvider.getMetadataTxnContext());
        for (Map.Entry<OperatorDescriptorId, IOperatorDescriptor> entry : operatorsMap.entrySet()) {
            IOperatorDescriptor opDesc = entry.getValue();
            OperatorDescriptorId oldId = opDesc.getOperatorId();
            OperatorDescriptorId opId = null;
            if (opDesc instanceof LSMTreeInsertDeleteOperatorDescriptor && ((LSMTreeInsertDeleteOperatorDescriptor) opDesc).isPrimary()) {
                metaOp = new FeedMetaOperatorDescriptor(jobSpec, feedConnectionId, opDesc, feedPolicyEntity.getProperties(), FeedRuntimeType.STORE);
                opId = metaOp.getOperatorId();
                opDesc.setOperatorId(opId);
            } else {
                if (opDesc instanceof AlgebricksMetaOperatorDescriptor) {
                    AlgebricksMetaOperatorDescriptor algOp = (AlgebricksMetaOperatorDescriptor) opDesc;
                    IPushRuntimeFactory[] runtimeFactories = algOp.getPipeline().getRuntimeFactories();
                    // Tweak AssignOp to work with messages
                    if (runtimeFactories[0] instanceof AssignRuntimeFactory && runtimeFactories.length > 1) {
                        IConnectorDescriptor connectorDesc = subJob.getOperatorInputMap().get(opDesc.getOperatorId()).get(0);
                        // anything on the network interface needs to be message compatible
                        if (connectorDesc instanceof MToNPartitioningConnectorDescriptor) {
                            metaOp = new FeedMetaOperatorDescriptor(jobSpec, feedConnectionId, opDesc, feedPolicyEntity.getProperties(), FeedRuntimeType.COMPUTE);
                            opId = metaOp.getOperatorId();
                            opDesc.setOperatorId(opId);
                        }
                    }
                }
                if (opId == null) {
                    opId = jobSpec.createOperatorDescriptorId(opDesc);
                }
            }
            operatorIdMapping.put(oldId, opId);
        }
        // copy connectors
        connectorIdMapping.clear();
        for (Entry<ConnectorDescriptorId, IConnectorDescriptor> entry : subJob.getConnectorMap().entrySet()) {
            IConnectorDescriptor connDesc = entry.getValue();
            ConnectorDescriptorId newConnId;
            if (connDesc instanceof MToNPartitioningConnectorDescriptor) {
                MToNPartitioningConnectorDescriptor m2nConn = (MToNPartitioningConnectorDescriptor) connDesc;
                connDesc = new MToNPartitioningWithMessageConnectorDescriptor(jobSpec, m2nConn.getTuplePartitionComputerFactory());
                newConnId = connDesc.getConnectorId();
            } else {
                newConnId = jobSpec.createConnectorDescriptor(connDesc);
            }
            connectorIdMapping.put(entry.getKey(), newConnId);
        }
        // make connections between operators
        for (Entry<ConnectorDescriptorId, Pair<Pair<IOperatorDescriptor, Integer>, Pair<IOperatorDescriptor, Integer>>> entry : subJob.getConnectorOperatorMap().entrySet()) {
            ConnectorDescriptorId newId = connectorIdMapping.get(entry.getKey());
            IConnectorDescriptor connDesc = jobSpec.getConnectorMap().get(newId);
            Pair<IOperatorDescriptor, Integer> leftOp = entry.getValue().getLeft();
            Pair<IOperatorDescriptor, Integer> rightOp = entry.getValue().getRight();
            IOperatorDescriptor leftOpDesc = jobSpec.getOperatorMap().get(leftOp.getLeft().getOperatorId());
            IOperatorDescriptor rightOpDesc = jobSpec.getOperatorMap().get(rightOp.getLeft().getOperatorId());
            if (leftOp.getLeft() instanceof FeedCollectOperatorDescriptor) {
                jobSpec.connect(new OneToOneConnectorDescriptor(jobSpec), replicateOp, iter1, leftOpDesc, leftOp.getRight());
            }
            jobSpec.connect(connDesc, leftOpDesc, leftOp.getRight(), rightOpDesc, rightOp.getRight());
        }
        // prepare for setting partition constraints
        operatorLocations.clear();
        operatorCounts.clear();
        for (Constraint constraint : subJob.getUserConstraints()) {
            LValueConstraintExpression lexpr = constraint.getLValue();
            ConstraintExpression cexpr = constraint.getRValue();
            OperatorDescriptorId opId;
            switch(lexpr.getTag()) {
                case PARTITION_COUNT:
                    opId = ((PartitionCountExpression) lexpr).getOperatorDescriptorId();
                    operatorCounts.put(operatorIdMapping.get(opId), (int) ((ConstantExpression) cexpr).getValue());
                    break;
                case PARTITION_LOCATION:
                    opId = ((PartitionLocationExpression) lexpr).getOperatorDescriptorId();
                    IOperatorDescriptor opDesc = jobSpec.getOperatorMap().get(operatorIdMapping.get(opId));
                    List<LocationConstraint> locations = operatorLocations.get(opDesc.getOperatorId());
                    if (locations == null) {
                        locations = new ArrayList<>();
                        operatorLocations.put(opDesc.getOperatorId(), locations);
                    }
                    String location = (String) ((ConstantExpression) cexpr).getValue();
                    LocationConstraint lc = new LocationConstraint(location, ((PartitionLocationExpression) lexpr).getPartition());
                    locations.add(lc);
                    break;
                default:
                    break;
            }
        }
        // set absolute location constraints
        for (Entry<OperatorDescriptorId, List<LocationConstraint>> entry : operatorLocations.entrySet()) {
            IOperatorDescriptor opDesc = jobSpec.getOperatorMap().get(entry.getKey());
            // why do we need to sort?
            Collections.sort(entry.getValue(), (LocationConstraint o1, LocationConstraint o2) -> {
                return o1.partition - o2.partition;
            });
            String[] locations = new String[entry.getValue().size()];
            for (int j = 0; j < locations.length; ++j) {
                locations[j] = entry.getValue().get(j).location;
            }
            PartitionConstraintHelper.addAbsoluteLocationConstraint(jobSpec, opDesc, locations);
        }
        // set count constraints
        for (Entry<OperatorDescriptorId, Integer> entry : operatorCounts.entrySet()) {
            IOperatorDescriptor opDesc = jobSpec.getOperatorMap().get(entry.getKey());
            if (!operatorLocations.keySet().contains(entry.getKey())) {
                PartitionConstraintHelper.addPartitionCountConstraint(jobSpec, opDesc, entry.getValue());
            }
        }
        // roots
        for (OperatorDescriptorId root : subJob.getRoots()) {
            jobSpec.addRoot(jobSpec.getOperatorMap().get(operatorIdMapping.get(root)));
        }
        jobIds.add(((JobEventListenerFactory) subJob.getJobletEventListenerFactory()).getJobId());
    }
    // jobEventListenerFactory
    jobSpec.setJobletEventListenerFactory(new MultiTransactionJobletEventListenerFactory(jobIds, true));
    // useConnectorSchedulingPolicy
    jobSpec.setUseConnectorPolicyForScheduling(jobsList.get(0).isUseConnectorPolicyForScheduling());
    // connectorAssignmentPolicy
    jobSpec.setConnectorPolicyAssignmentPolicy(jobsList.get(0).getConnectorPolicyAssignmentPolicy());
    return jobSpec;
}
Also used : HashMap(java.util.HashMap) AlgebricksPartitionConstraint(org.apache.hyracks.algebricks.common.constraints.AlgebricksPartitionConstraint) AlgebricksAbsolutePartitionConstraint(org.apache.hyracks.algebricks.common.constraints.AlgebricksAbsolutePartitionConstraint) Constraint(org.apache.hyracks.api.constraints.Constraint) LocationConstraint(org.apache.asterix.metadata.feeds.LocationConstraint) ConstantExpression(org.apache.hyracks.api.constraints.expressions.ConstantExpression) ConnectorDescriptorId(org.apache.hyracks.api.dataflow.ConnectorDescriptorId) ArrayList(java.util.ArrayList) OneToOneConnectorDescriptor(org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor) LocationConstraint(org.apache.asterix.metadata.feeds.LocationConstraint) List(java.util.List) ArrayList(java.util.ArrayList) AlgebricksMetaOperatorDescriptor(org.apache.hyracks.algebricks.runtime.operators.meta.AlgebricksMetaOperatorDescriptor) LValueConstraintExpression(org.apache.hyracks.api.constraints.expressions.LValueConstraintExpression) ReplicateOperatorDescriptor(org.apache.hyracks.dataflow.std.misc.ReplicateOperatorDescriptor) Map(java.util.Map) HashMap(java.util.HashMap) FeedCollectOperatorDescriptor(org.apache.asterix.external.operators.FeedCollectOperatorDescriptor) FeedPolicyEntity(org.apache.asterix.metadata.entities.FeedPolicyEntity) JobSpecification(org.apache.hyracks.api.job.JobSpecification) FeedIntakeOperatorDescriptor(org.apache.asterix.external.operators.FeedIntakeOperatorDescriptor) JobId(org.apache.asterix.common.transactions.JobId) Pair(org.apache.commons.lang3.tuple.Pair) IConnectorDescriptor(org.apache.hyracks.api.dataflow.IConnectorDescriptor) MToNPartitioningWithMessageConnectorDescriptor(org.apache.hyracks.dataflow.std.connectors.MToNPartitioningWithMessageConnectorDescriptor) OperatorDescriptorId(org.apache.hyracks.api.dataflow.OperatorDescriptorId) LValueConstraintExpression(org.apache.hyracks.api.constraints.expressions.LValueConstraintExpression) ConstraintExpression(org.apache.hyracks.api.constraints.expressions.ConstraintExpression) FeedConnection(org.apache.asterix.metadata.entities.FeedConnection) MultiTransactionJobletEventListenerFactory(org.apache.asterix.runtime.job.listener.MultiTransactionJobletEventListenerFactory) MToNPartitioningConnectorDescriptor(org.apache.hyracks.dataflow.std.connectors.MToNPartitioningConnectorDescriptor) AssignRuntimeFactory(org.apache.hyracks.algebricks.runtime.operators.std.AssignRuntimeFactory) IPushRuntimeFactory(org.apache.hyracks.algebricks.runtime.base.IPushRuntimeFactory) AlgebricksPartitionConstraint(org.apache.hyracks.algebricks.common.constraints.AlgebricksPartitionConstraint) AlgebricksAbsolutePartitionConstraint(org.apache.hyracks.algebricks.common.constraints.AlgebricksAbsolutePartitionConstraint) Constraint(org.apache.hyracks.api.constraints.Constraint) LocationConstraint(org.apache.asterix.metadata.feeds.LocationConstraint) FeedMetaOperatorDescriptor(org.apache.asterix.external.operators.FeedMetaOperatorDescriptor) IOperatorDescriptor(org.apache.hyracks.api.dataflow.IOperatorDescriptor) FeedConnectionId(org.apache.asterix.external.feed.management.FeedConnectionId) LSMTreeInsertDeleteOperatorDescriptor(org.apache.asterix.common.dataflow.LSMTreeInsertDeleteOperatorDescriptor)

Example 2 with FeedIntakeOperatorDescriptor

use of org.apache.asterix.external.operators.FeedIntakeOperatorDescriptor in project asterixdb by apache.

the class MetadataProvider method buildFeedIntakeRuntime.

public Triple<IOperatorDescriptor, AlgebricksPartitionConstraint, IAdapterFactory> buildFeedIntakeRuntime(JobSpecification jobSpec, Feed primaryFeed, FeedPolicyAccessor policyAccessor) throws Exception {
    Triple<IAdapterFactory, RecordDescriptor, IDataSourceAdapter.AdapterType> factoryOutput;
    factoryOutput = FeedMetadataUtil.getPrimaryFeedFactoryAndOutput(primaryFeed, policyAccessor, mdTxnCtx, getApplicationContext());
    ARecordType recordType = FeedMetadataUtil.getOutputType(primaryFeed, primaryFeed.getAdapterConfiguration(), ExternalDataConstants.KEY_TYPE_NAME);
    IAdapterFactory adapterFactory = factoryOutput.first;
    FeedIntakeOperatorDescriptor feedIngestor = null;
    switch(factoryOutput.third) {
        case INTERNAL:
            feedIngestor = new FeedIntakeOperatorDescriptor(jobSpec, primaryFeed, adapterFactory, recordType, policyAccessor, factoryOutput.second);
            break;
        case EXTERNAL:
            String libraryName = primaryFeed.getAdapterName().trim().split(FeedConstants.NamingConstants.LIBRARY_NAME_SEPARATOR)[0];
            feedIngestor = new FeedIntakeOperatorDescriptor(jobSpec, primaryFeed, libraryName, adapterFactory.getClass().getName(), recordType, policyAccessor, factoryOutput.second);
            break;
        default:
            break;
    }
    AlgebricksPartitionConstraint partitionConstraint = adapterFactory.getPartitionConstraint();
    return new Triple<>(feedIngestor, partitionConstraint, adapterFactory);
}
Also used : Triple(org.apache.hyracks.algebricks.common.utils.Triple) RecordDescriptor(org.apache.hyracks.api.dataflow.value.RecordDescriptor) IAdapterFactory(org.apache.asterix.external.api.IAdapterFactory) AlgebricksPartitionConstraint(org.apache.hyracks.algebricks.common.constraints.AlgebricksPartitionConstraint) ARecordType(org.apache.asterix.om.types.ARecordType) FeedIntakeOperatorDescriptor(org.apache.asterix.external.operators.FeedIntakeOperatorDescriptor)

Aggregations

FeedIntakeOperatorDescriptor (org.apache.asterix.external.operators.FeedIntakeOperatorDescriptor)2 AlgebricksPartitionConstraint (org.apache.hyracks.algebricks.common.constraints.AlgebricksPartitionConstraint)2 ArrayList (java.util.ArrayList)1 HashMap (java.util.HashMap)1 List (java.util.List)1 Map (java.util.Map)1 LSMTreeInsertDeleteOperatorDescriptor (org.apache.asterix.common.dataflow.LSMTreeInsertDeleteOperatorDescriptor)1 JobId (org.apache.asterix.common.transactions.JobId)1 IAdapterFactory (org.apache.asterix.external.api.IAdapterFactory)1 FeedConnectionId (org.apache.asterix.external.feed.management.FeedConnectionId)1 FeedCollectOperatorDescriptor (org.apache.asterix.external.operators.FeedCollectOperatorDescriptor)1 FeedMetaOperatorDescriptor (org.apache.asterix.external.operators.FeedMetaOperatorDescriptor)1 FeedConnection (org.apache.asterix.metadata.entities.FeedConnection)1 FeedPolicyEntity (org.apache.asterix.metadata.entities.FeedPolicyEntity)1 LocationConstraint (org.apache.asterix.metadata.feeds.LocationConstraint)1 ARecordType (org.apache.asterix.om.types.ARecordType)1 MultiTransactionJobletEventListenerFactory (org.apache.asterix.runtime.job.listener.MultiTransactionJobletEventListenerFactory)1 Pair (org.apache.commons.lang3.tuple.Pair)1 AlgebricksAbsolutePartitionConstraint (org.apache.hyracks.algebricks.common.constraints.AlgebricksAbsolutePartitionConstraint)1 Triple (org.apache.hyracks.algebricks.common.utils.Triple)1