Examples with LocalDataSegmentPuller org.apache.druid.segment.loading.LocalDataSegmentPuller used on opensource projects

Example 1 with LocalDataSegmentPuller

use of org.apache.druid.segment.loading.LocalDataSegmentPuller in project druid by druid-io.

the class SegmentManagerBroadcastJoinIndexedTableTest method setup.

@Before
public void setup() throws IOException {
    segmentPuller = new LocalDataSegmentPuller();
    objectMapper = new DefaultObjectMapper().registerModule(new SegmentizerModule()).registerModule(new SimpleModule().registerSubtypes(new NamedType(LocalLoadSpec.class, "local")));
    indexIO = new IndexIO(objectMapper, () -> 0);
    objectMapper.setInjectableValues(new InjectableValues.Std().addValue(LocalDataSegmentPuller.class, segmentPuller).addValue(ExprMacroTable.class.getName(), TestExprMacroTable.INSTANCE).addValue(ObjectMapper.class.getName(), objectMapper).addValue(IndexIO.class, indexIO));
    segmentCacheDir = temporaryFolder.newFolder();
    segmentDeepStorageDir = temporaryFolder.newFolder();
    segmentCacheManager = new SegmentLocalCacheManager(new SegmentLoaderConfig() {

        @Override
        public List<StorageLocationConfig> getLocations() {
            return Collections.singletonList(new StorageLocationConfig(segmentCacheDir, null, null));
        }
    }, objectMapper);
    segmentManager = new SegmentManager(new SegmentLocalCacheLoader(segmentCacheManager, indexIO, objectMapper));
    joinableFactory = new BroadcastTableJoinableFactory(segmentManager);
    EmittingLogger.registerEmitter(new NoopServiceEmitter());
}

Example 2 with LocalDataSegmentPuller

use of org.apache.druid.segment.loading.LocalDataSegmentPuller in project hive by apache.

the class TestDruidRecordWriter method testWrite.

// Test is failing due to Guava dependency, Druid 0.13.0 should have less dependency on Guava
@Ignore
@Test
public void testWrite() throws IOException, SegmentLoadingException {
    final String dataSourceName = "testDataSource";
    final File segmentOutputDir = temporaryFolder.newFolder();
    final File workingDir = temporaryFolder.newFolder();
    Configuration config = new Configuration();
    final InputRowParser inputRowParser = new MapInputRowParser(new TimeAndDimsParseSpec(new TimestampSpec(DruidConstants.DEFAULT_TIMESTAMP_COLUMN, "auto", null), new DimensionsSpec(ImmutableList.of(new StringDimensionSchema("host")), null, null)));
    final Map<String, Object> parserMap = objectMapper.convertValue(inputRowParser, new TypeReference<Map<String, Object>>() {
    });
    DataSchema dataSchema = new DataSchema(dataSourceName, parserMap, new AggregatorFactory[] { new LongSumAggregatorFactory("visited_sum", "visited_sum"), new HyperUniquesAggregatorFactory("unique_hosts", "unique_hosts") }, new UniformGranularitySpec(Granularities.DAY, Granularities.NONE, ImmutableList.of(INTERVAL_FULL)), null, objectMapper);
    IndexSpec indexSpec = new IndexSpec(new RoaringBitmapSerdeFactory(true), null, null, null);
    RealtimeTuningConfig tuningConfig = new RealtimeTuningConfig(null, null, null, null, temporaryFolder.newFolder(), null, null, null, null, indexSpec, null, null, 0, 0, null, null, 0L, null, null);
    LocalFileSystem localFileSystem = FileSystem.getLocal(config);
    DataSegmentPusher dataSegmentPusher = new LocalDataSegmentPusher(new LocalDataSegmentPusherConfig() {

        @Override
        public File getStorageDirectory() {
            return segmentOutputDir;
        }
    });
    Path segmentDescriptorPath = new Path(workingDir.getAbsolutePath(), DruidStorageHandler.SEGMENTS_DESCRIPTOR_DIR_NAME);
    DruidRecordWriter druidRecordWriter = new DruidRecordWriter(dataSchema, tuningConfig, dataSegmentPusher, 20, segmentDescriptorPath, localFileSystem);
    List<DruidWritable> druidWritables = expectedRows.stream().map(input -> new DruidWritable(ImmutableMap.<String, Object>builder().putAll(input).put(Constants.DRUID_TIMESTAMP_GRANULARITY_COL_NAME, Granularities.DAY.bucketStart(new DateTime((long) input.get(DruidConstants.DEFAULT_TIMESTAMP_COLUMN))).getMillis()).build())).collect(Collectors.toList());
    for (DruidWritable druidWritable : druidWritables) {
        druidRecordWriter.write(druidWritable);
    }
    druidRecordWriter.close(false);
    List<DataSegment> dataSegmentList = DruidStorageHandlerUtils.getCreatedSegments(segmentDescriptorPath, config);
    Assert.assertEquals(1, dataSegmentList.size());
    File tmpUnzippedSegmentDir = temporaryFolder.newFolder();
    new LocalDataSegmentPuller().getSegmentFiles(dataSegmentList.get(0), tmpUnzippedSegmentDir);
    final QueryableIndex queryableIndex = DruidStorageHandlerUtils.INDEX_IO.loadIndex(tmpUnzippedSegmentDir);
    QueryableIndexStorageAdapter adapter = new QueryableIndexStorageAdapter(queryableIndex);
    Firehose firehose = new IngestSegmentFirehose(ImmutableList.of(new WindowedStorageAdapter(adapter, adapter.getInterval())), null, ImmutableList.of("host"), ImmutableList.of("visited_sum", "unique_hosts"), null);
    List<InputRow> rows = Lists.newArrayList();
    while (firehose.hasMore()) {
        rows.add(firehose.nextRow());
    }
    verifyRows(expectedRows, rows);
}

Example 3 with LocalDataSegmentPuller

use of org.apache.druid.segment.loading.LocalDataSegmentPuller in project druid by druid-io.

the class AbstractParallelIndexSupervisorTaskTest method prepareObjectMapper.

public void prepareObjectMapper(ObjectMapper objectMapper, IndexIO indexIO) {
    final TaskConfig taskConfig = new TaskConfig(null, null, null, null, null, false, null, null, null, false, false, TaskConfig.BATCH_PROCESSING_MODE_DEFAULT.name());
    objectMapper.setInjectableValues(new InjectableValues.Std().addValue(ExprMacroTable.class, LookupEnabledTestExprMacroTable.INSTANCE).addValue(IndexIO.class, indexIO).addValue(ObjectMapper.class, objectMapper).addValue(ChatHandlerProvider.class, new NoopChatHandlerProvider()).addValue(AuthConfig.class, new AuthConfig()).addValue(AuthorizerMapper.class, null).addValue(RowIngestionMetersFactory.class, new DropwizardRowIngestionMetersFactory()).addValue(DataSegment.PruneSpecsHolder.class, DataSegment.PruneSpecsHolder.DEFAULT).addValue(AuthorizerMapper.class, new AuthorizerMapper(ImmutableMap.of())).addValue(AppenderatorsManager.class, TestUtils.APPENDERATORS_MANAGER).addValue(LocalDataSegmentPuller.class, new LocalDataSegmentPuller()).addValue(CoordinatorClient.class, coordinatorClient).addValue(SegmentCacheManagerFactory.class, new SegmentCacheManagerFactory(objectMapper)).addValue(RetryPolicyFactory.class, new RetryPolicyFactory(new RetryPolicyConfig())).addValue(TaskConfig.class, taskConfig));
    objectMapper.registerSubtypes(new NamedType(ParallelIndexSupervisorTask.class, ParallelIndexSupervisorTask.TYPE), new NamedType(CompactionTask.CompactionTuningConfig.class, CompactionTask.CompactionTuningConfig.TYPE), new NamedType(SinglePhaseSubTask.class, SinglePhaseSubTask.TYPE), new NamedType(PartialHashSegmentGenerateTask.class, PartialHashSegmentGenerateTask.TYPE), new NamedType(PartialRangeSegmentGenerateTask.class, PartialRangeSegmentGenerateTask.TYPE), new NamedType(PartialGenericSegmentMergeTask.class, PartialGenericSegmentMergeTask.TYPE), new NamedType(PartialDimensionDistributionTask.class, PartialDimensionDistributionTask.TYPE), new NamedType(PartialDimensionCardinalityTask.class, PartialDimensionCardinalityTask.TYPE));
}

Example 4 with LocalDataSegmentPuller

use of org.apache.druid.segment.loading.LocalDataSegmentPuller in project druid by druid-io.

the class BatchDeltaIngestionTest method testIngestion.

private void testIngestion(HadoopDruidIndexerConfig config, List<ImmutableMap<String, Object>> expectedRowsGenerated, WindowedDataSegment windowedDataSegment, List<String> expectedDimensions, List<String> expectedMetrics) throws Exception {
    IndexGeneratorJob job = new IndexGeneratorJob(config);
    Assert.assertTrue(JobHelper.runJobs(ImmutableList.of(job)));
    List<DataSegmentAndIndexZipFilePath> dataSegmentAndIndexZipFilePaths = IndexGeneratorJob.getPublishedSegmentAndIndexZipFilePaths(config);
    JobHelper.renameIndexFilesForSegments(config.getSchema(), dataSegmentAndIndexZipFilePaths);
    JobHelper.maybeDeleteIntermediatePath(true, config.getSchema());
    File workingPath = new File(config.makeIntermediatePath().toUri().getPath());
    Assert.assertFalse(workingPath.exists());
    File segmentFolder = new File(StringUtils.format("%s/%s/%s_%s/%s/0", config.getSchema().getIOConfig().getSegmentOutputPath(), config.getSchema().getDataSchema().getDataSource(), INTERVAL_FULL.getStart().toString(), INTERVAL_FULL.getEnd().toString(), config.getSchema().getTuningConfig().getVersion()));
    Assert.assertTrue(segmentFolder.exists());
    File indexZip = new File(segmentFolder, "index.zip");
    Assert.assertTrue(indexZip.exists());
    File tmpUnzippedSegmentDir = temporaryFolder.newFolder();
    new LocalDataSegmentPuller().getSegmentFiles(indexZip, tmpUnzippedSegmentDir);
    QueryableIndex index = INDEX_IO.loadIndex(tmpUnzippedSegmentDir);
    StorageAdapter adapter = new QueryableIndexStorageAdapter(index);
    Firehose firehose = new IngestSegmentFirehose(ImmutableList.of(new WindowedStorageAdapter(adapter, windowedDataSegment.getInterval())), TransformSpec.NONE, expectedDimensions, expectedMetrics, null);
    List<InputRow> rows = new ArrayList<>();
    while (firehose.hasMore()) {
        rows.add(firehose.nextRow());
    }
    verifyRows(expectedRowsGenerated, rows, expectedDimensions, expectedMetrics);
}