Examples with HoodieClusteringGroup org.apache.hudi.avro.model.HoodieClusteringGroup used on opensource projects

Example 6 with HoodieClusteringGroup

use of org.apache.hudi.avro.model.HoodieClusteringGroup in project hudi by apache.

the class TestSimpleConcurrentFileWritesConflictResolutionStrategy method createReplaceRequested.

private void createReplaceRequested(String instantTime) throws Exception {
    String fileId1 = "file-1";
    String fileId2 = "file-2";
    // create replace instant to mark fileId1 as deleted
    HoodieRequestedReplaceMetadata requestedReplaceMetadata = new HoodieRequestedReplaceMetadata();
    requestedReplaceMetadata.setOperationType(WriteOperationType.CLUSTER.name());
    HoodieClusteringPlan clusteringPlan = new HoodieClusteringPlan();
    HoodieClusteringGroup clusteringGroup = new HoodieClusteringGroup();
    HoodieSliceInfo sliceInfo = new HoodieSliceInfo();
    sliceInfo.setFileId(fileId1);
    sliceInfo.setPartitionPath(HoodieTestDataGenerator.DEFAULT_FIRST_PARTITION_PATH);
    clusteringGroup.setSlices(Arrays.asList(sliceInfo));
    clusteringPlan.setInputGroups(Arrays.asList(clusteringGroup));
    requestedReplaceMetadata.setClusteringPlan(clusteringPlan);
    requestedReplaceMetadata.setVersion(TimelineLayoutVersion.CURR_VERSION);
    HoodieTestTable.of(metaClient).addRequestedReplace(instantTime, Option.of(requestedReplaceMetadata)).withBaseFilesInPartition(HoodieTestDataGenerator.DEFAULT_FIRST_PARTITION_PATH, fileId1, fileId2);
}

Example 7 with HoodieClusteringGroup

use of org.apache.hudi.avro.model.HoodieClusteringGroup in project hudi by apache.

the class ClusteringUtils method createClusteringPlan.

/**
 * Create clustering plan from input fileSliceGroups.
 */
public static HoodieClusteringPlan createClusteringPlan(String strategyClassName, Map<String, String> strategyParams, List<FileSlice>[] fileSliceGroups, Map<String, String> extraMetadata) {
    List<HoodieClusteringGroup> clusteringGroups = Arrays.stream(fileSliceGroups).map(fileSliceGroup -> {
        Map<String, Double> groupMetrics = buildMetrics(fileSliceGroup);
        List<HoodieSliceInfo> sliceInfos = getFileSliceInfo(fileSliceGroup);
        return HoodieClusteringGroup.newBuilder().setSlices(sliceInfos).setMetrics(groupMetrics).build();
    }).collect(Collectors.toList());
    HoodieClusteringStrategy strategy = HoodieClusteringStrategy.newBuilder().setStrategyClassName(strategyClassName).setStrategyParams(strategyParams).build();
    return HoodieClusteringPlan.newBuilder().setInputGroups(clusteringGroups).setExtraMetadata(extraMetadata).setStrategy(strategy).build();
}

Example 8 with HoodieClusteringGroup

use of org.apache.hudi.avro.model.HoodieClusteringGroup in project hudi by apache.

the class JavaExecutionStrategy method runClusteringForGroup.

/**
 * Executes clustering for the group.
 */
private List<WriteStatus> runClusteringForGroup(HoodieClusteringGroup clusteringGroup, Map<String, String> strategyParams, boolean preserveHoodieMetadata, String instantTime) {
    List<HoodieRecord<T>> inputRecords = readRecordsForGroup(clusteringGroup, instantTime);
    Schema readerSchema = HoodieAvroUtils.addMetadataFields(new Schema.Parser().parse(getWriteConfig().getSchema()));
    List<HoodieFileGroupId> inputFileIds = clusteringGroup.getSlices().stream().map(info -> new HoodieFileGroupId(info.getPartitionPath(), info.getFileId())).collect(Collectors.toList());
    return performClusteringWithRecordList(inputRecords, clusteringGroup.getNumOutputFileGroups(), instantTime, strategyParams, readerSchema, inputFileIds, preserveHoodieMetadata);
}