use of org.apache.hyracks.dataflow.std.file.IFileSplitProvider in project asterixdb by apache.
the class ExternalIndexingOperations method buildFilesIndexCreateJobSpec.
public static JobSpecification buildFilesIndexCreateJobSpec(Dataset dataset, List<ExternalFile> externalFilesSnapshot, MetadataProvider metadataProvider) throws AlgebricksException {
IStorageComponentProvider storageComponentProvider = metadataProvider.getStorageComponentProvider();
JobSpecification spec = RuntimeUtils.createJobSpecification(metadataProvider.getApplicationContext());
Pair<ILSMMergePolicyFactory, Map<String, String>> compactionInfo = DatasetUtil.getMergePolicyFactory(dataset, metadataProvider.getMetadataTxnContext());
ILSMMergePolicyFactory mergePolicyFactory = compactionInfo.first;
Map<String, String> mergePolicyProperties = compactionInfo.second;
Pair<IFileSplitProvider, AlgebricksPartitionConstraint> secondarySplitsAndConstraint = metadataProvider.getSplitProviderAndConstraints(dataset, IndexingConstants.getFilesIndexName(dataset.getDatasetName()));
IFileSplitProvider secondaryFileSplitProvider = secondarySplitsAndConstraint.first;
String fileIndexName = IndexingConstants.getFilesIndexName(dataset.getDatasetName());
Index fileIndex = MetadataManager.INSTANCE.getIndex(metadataProvider.getMetadataTxnContext(), dataset.getDataverseName(), dataset.getDatasetName(), fileIndexName);
ARecordType recordType = (ARecordType) metadataProvider.findType(dataset.getItemTypeDataverseName(), dataset.getItemTypeName());
IResourceFactory resourceFactory = dataset.getResourceFactory(metadataProvider, fileIndex, recordType, null, mergePolicyFactory, mergePolicyProperties);
IIndexBuilderFactory indexBuilderFactory = new IndexBuilderFactory(storageComponentProvider.getStorageManager(), secondaryFileSplitProvider, resourceFactory, !dataset.isTemp());
IIndexDataflowHelperFactory dataflowHelperFactory = new IndexDataflowHelperFactory(storageComponentProvider.getStorageManager(), secondaryFileSplitProvider);
ExternalFilesIndexCreateOperatorDescriptor externalFilesOp = new ExternalFilesIndexCreateOperatorDescriptor(spec, indexBuilderFactory, dataflowHelperFactory, externalFilesSnapshot);
AlgebricksPartitionConstraintHelper.setPartitionConstraintInJobSpec(spec, externalFilesOp, secondarySplitsAndConstraint.second);
spec.addRoot(externalFilesOp);
spec.setConnectorPolicyAssignmentPolicy(new ConnectorPolicyAssignmentPolicy());
return spec;
}
use of org.apache.hyracks.dataflow.std.file.IFileSplitProvider in project asterixdb by apache.
the class ExternalIndexingOperations method buildDropFilesIndexJobSpec.
public static JobSpecification buildDropFilesIndexJobSpec(MetadataProvider metadataProvider, Dataset dataset) throws AlgebricksException {
String indexName = IndexingConstants.getFilesIndexName(dataset.getDatasetName());
JobSpecification spec = RuntimeUtils.createJobSpecification(metadataProvider.getApplicationContext());
Pair<IFileSplitProvider, AlgebricksPartitionConstraint> splitsAndConstraint = metadataProvider.getSplitProviderAndConstraints(dataset, indexName);
IIndexDataflowHelperFactory dataflowHelperFactory = new IndexDataflowHelperFactory(metadataProvider.getStorageComponentProvider().getStorageManager(), splitsAndConstraint.first);
IndexDropOperatorDescriptor btreeDrop = new IndexDropOperatorDescriptor(spec, dataflowHelperFactory);
AlgebricksPartitionConstraintHelper.setPartitionConstraintInJobSpec(spec, btreeDrop, splitsAndConstraint.second);
spec.addRoot(btreeDrop);
return spec;
}
Aggregations