Search in sources :

Example 36 with QueryableIndex

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

the class SqlExpressionBenchmark method setup.

@Setup(Level.Trial)
public void setup() {
    final GeneratorSchemaInfo schemaInfo = GeneratorBasicSchemas.SCHEMA_MAP.get("expression-testbench");
    final DataSegment dataSegment = DataSegment.builder().dataSource("foo").interval(schemaInfo.getDataInterval()).version("1").shardSpec(new LinearShardSpec(0)).size(0).build();
    final PlannerConfig plannerConfig = new PlannerConfig();
    final SegmentGenerator segmentGenerator = closer.register(new SegmentGenerator());
    log.info("Starting benchmark setup using cacheDir[%s], rows[%,d].", segmentGenerator.getCacheDir(), rowsPerSegment);
    final QueryableIndex index = segmentGenerator.generate(dataSegment, schemaInfo, Granularities.NONE, rowsPerSegment);
    final QueryRunnerFactoryConglomerate conglomerate = QueryStackTests.createQueryRunnerFactoryConglomerate(closer, PROCESSING_CONFIG);
    final SpecificSegmentsQuerySegmentWalker walker = new SpecificSegmentsQuerySegmentWalker(conglomerate).add(dataSegment, index);
    closer.register(walker);
    final DruidSchemaCatalog rootSchema = CalciteTests.createMockRootSchema(conglomerate, walker, plannerConfig, AuthTestUtils.TEST_AUTHORIZER_MAPPER);
    plannerFactory = new PlannerFactory(rootSchema, CalciteTests.createMockQueryMakerFactory(walker, conglomerate), CalciteTests.createOperatorTable(), CalciteTests.createExprMacroTable(), plannerConfig, AuthTestUtils.TEST_AUTHORIZER_MAPPER, CalciteTests.getJsonMapper(), CalciteTests.DRUID_SCHEMA_NAME);
    try {
        SqlVectorizedExpressionSanityTest.sanityTestVectorizedSqlQueries(plannerFactory, QUERIES.get(Integer.parseInt(query)));
    } catch (Throwable ignored) {
    // the show must go on
    }
}
Also used : SegmentGenerator(org.apache.druid.segment.generator.SegmentGenerator) QueryRunnerFactoryConglomerate(org.apache.druid.query.QueryRunnerFactoryConglomerate) SpecificSegmentsQuerySegmentWalker(org.apache.druid.sql.calcite.util.SpecificSegmentsQuerySegmentWalker) LinearShardSpec(org.apache.druid.timeline.partition.LinearShardSpec) QueryableIndex(org.apache.druid.segment.QueryableIndex) GeneratorSchemaInfo(org.apache.druid.segment.generator.GeneratorSchemaInfo) PlannerConfig(org.apache.druid.sql.calcite.planner.PlannerConfig) DruidSchemaCatalog(org.apache.druid.sql.calcite.schema.DruidSchemaCatalog) PlannerFactory(org.apache.druid.sql.calcite.planner.PlannerFactory) DataSegment(org.apache.druid.timeline.DataSegment) Setup(org.openjdk.jmh.annotations.Setup)

Example 37 with QueryableIndex

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

the class TopNTypeInterfaceBenchmark method setup.

@Setup
public void setup() throws IOException {
    log.info("SETUP CALLED AT " + System.currentTimeMillis());
    ComplexMetrics.registerSerde("hyperUnique", new HyperUniquesSerde());
    setupQueries();
    schemaInfo = GeneratorBasicSchemas.SCHEMA_MAP.get("basic");
    queryBuilder = SCHEMA_QUERY_MAP.get("basic").get("string");
    queryBuilder.threshold(threshold);
    stringQuery = queryBuilder.build();
    TopNQueryBuilder longBuilder = SCHEMA_QUERY_MAP.get("basic").get("long");
    longBuilder.threshold(threshold);
    longQuery = longBuilder.build();
    TopNQueryBuilder floatBuilder = SCHEMA_QUERY_MAP.get("basic").get("float");
    floatBuilder.threshold(threshold);
    floatQuery = floatBuilder.build();
    incIndexes = new ArrayList<>();
    for (int i = 0; i < numSegments; i++) {
        log.info("Generating rows for segment " + i);
        DataGenerator gen = new DataGenerator(schemaInfo.getColumnSchemas(), RNG_SEED + i, schemaInfo.getDataInterval(), rowsPerSegment);
        IncrementalIndex incIndex = makeIncIndex();
        for (int j = 0; j < rowsPerSegment; j++) {
            InputRow row = gen.nextRow();
            if (j % 10000 == 0) {
                log.info(j + " rows generated.");
            }
            incIndex.add(row);
        }
        incIndexes.add(incIndex);
    }
    File tmpFile = FileUtils.createTempDir();
    log.info("Using temp dir: " + tmpFile.getAbsolutePath());
    tmpFile.deleteOnExit();
    qIndexes = new ArrayList<>();
    for (int i = 0; i < numSegments; i++) {
        File indexFile = INDEX_MERGER_V9.persist(incIndexes.get(i), tmpFile, new IndexSpec(), null);
        QueryableIndex qIndex = INDEX_IO.loadIndex(indexFile);
        qIndexes.add(qIndex);
    }
    factory = new TopNQueryRunnerFactory(new StupidPool<>("TopNBenchmark-compute-bufferPool", new OffheapBufferGenerator("compute", 250000000), 0, Integer.MAX_VALUE), new TopNQueryQueryToolChest(new TopNQueryConfig()), QueryBenchmarkUtil.NOOP_QUERYWATCHER);
}
Also used : TopNQueryBuilder(org.apache.druid.query.topn.TopNQueryBuilder) IndexSpec(org.apache.druid.segment.IndexSpec) IncrementalIndex(org.apache.druid.segment.incremental.IncrementalIndex) OnheapIncrementalIndex(org.apache.druid.segment.incremental.OnheapIncrementalIndex) HyperUniquesSerde(org.apache.druid.query.aggregation.hyperloglog.HyperUniquesSerde) OffheapBufferGenerator(org.apache.druid.offheap.OffheapBufferGenerator) TopNQueryConfig(org.apache.druid.query.topn.TopNQueryConfig) QueryableIndex(org.apache.druid.segment.QueryableIndex) DataGenerator(org.apache.druid.segment.generator.DataGenerator) InputRow(org.apache.druid.data.input.InputRow) TopNQueryRunnerFactory(org.apache.druid.query.topn.TopNQueryRunnerFactory) StupidPool(org.apache.druid.collections.StupidPool) TopNQueryQueryToolChest(org.apache.druid.query.topn.TopNQueryQueryToolChest) File(java.io.File) Setup(org.openjdk.jmh.annotations.Setup)

Example 38 with QueryableIndex

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

the class TimeCompareBenchmark method setup.

@Setup
public void setup() throws IOException {
    log.info("SETUP CALLED AT " + System.currentTimeMillis());
    ComplexMetrics.registerSerde("hyperUnique", new HyperUniquesSerde());
    executorService = Execs.multiThreaded(numSegments, "TopNThreadPool");
    setupQueries();
    String schemaName = "basic";
    schemaInfo = GeneratorBasicSchemas.SCHEMA_MAP.get(schemaName);
    segmentIntervals = new Interval[numSegments];
    long startMillis = schemaInfo.getDataInterval().getStartMillis();
    long endMillis = schemaInfo.getDataInterval().getEndMillis();
    long partialIntervalMillis = (endMillis - startMillis) / numSegments;
    for (int i = 0; i < numSegments; i++) {
        long partialEndMillis = startMillis + partialIntervalMillis;
        segmentIntervals[i] = Intervals.utc(startMillis, partialEndMillis);
        log.info("Segment [%d] with interval [%s]", i, segmentIntervals[i]);
        startMillis = partialEndMillis;
    }
    incIndexes = new ArrayList<>();
    for (int i = 0; i < numSegments; i++) {
        log.info("Generating rows for segment " + i);
        DataGenerator gen = new DataGenerator(schemaInfo.getColumnSchemas(), RNG_SEED + i, segmentIntervals[i], rowsPerSegment);
        IncrementalIndex incIndex = makeIncIndex();
        for (int j = 0; j < rowsPerSegment; j++) {
            InputRow row = gen.nextRow();
            if (j % 10000 == 0) {
                log.info(j + " rows generated.");
            }
            incIndex.add(row);
        }
        incIndexes.add(incIndex);
    }
    tmpDir = FileUtils.createTempDir();
    log.info("Using temp dir: " + tmpDir.getAbsolutePath());
    qIndexes = new ArrayList<>();
    for (int i = 0; i < numSegments; i++) {
        File indexFile = INDEX_MERGER_V9.persist(incIndexes.get(i), tmpDir, new IndexSpec(), null);
        QueryableIndex qIndex = INDEX_IO.loadIndex(indexFile);
        qIndexes.add(qIndex);
    }
    List<QueryRunner<Result<TopNResultValue>>> singleSegmentRunners = new ArrayList<>();
    QueryToolChest toolChest = topNFactory.getToolchest();
    for (int i = 0; i < numSegments; i++) {
        SegmentId segmentId = SegmentId.dummy("qIndex " + i);
        QueryRunner<Result<TopNResultValue>> runner = QueryBenchmarkUtil.makeQueryRunner(topNFactory, segmentId, new QueryableIndexSegment(qIndexes.get(i), segmentId));
        singleSegmentRunners.add(new PerSegmentOptimizingQueryRunner<>(toolChest.preMergeQueryDecoration(runner), new PerSegmentQueryOptimizationContext(new SegmentDescriptor(segmentIntervals[i], "1", 0))));
    }
    topNRunner = toolChest.postMergeQueryDecoration(new FinalizeResultsQueryRunner<>(toolChest.mergeResults(topNFactory.mergeRunners(executorService, singleSegmentRunners)), toolChest));
    List<QueryRunner<Result<TimeseriesResultValue>>> singleSegmentRunnersT = new ArrayList<>();
    QueryToolChest toolChestT = timeseriesFactory.getToolchest();
    for (int i = 0; i < numSegments; i++) {
        SegmentId segmentId = SegmentId.dummy("qIndex " + i);
        QueryRunner<Result<TimeseriesResultValue>> runner = QueryBenchmarkUtil.makeQueryRunner(timeseriesFactory, segmentId, new QueryableIndexSegment(qIndexes.get(i), segmentId));
        singleSegmentRunnersT.add(new PerSegmentOptimizingQueryRunner<>(toolChestT.preMergeQueryDecoration(runner), new PerSegmentQueryOptimizationContext(new SegmentDescriptor(segmentIntervals[i], "1", 0))));
    }
    timeseriesRunner = toolChestT.postMergeQueryDecoration(new FinalizeResultsQueryRunner<>(toolChestT.mergeResults(timeseriesFactory.mergeRunners(executorService, singleSegmentRunnersT)), toolChestT));
}
Also used : QueryableIndexSegment(org.apache.druid.segment.QueryableIndexSegment) TimeseriesResultValue(org.apache.druid.query.timeseries.TimeseriesResultValue) IndexSpec(org.apache.druid.segment.IndexSpec) ArrayList(java.util.ArrayList) HyperUniquesSerde(org.apache.druid.query.aggregation.hyperloglog.HyperUniquesSerde) TopNQueryQueryToolChest(org.apache.druid.query.topn.TopNQueryQueryToolChest) TimeseriesQueryQueryToolChest(org.apache.druid.query.timeseries.TimeseriesQueryQueryToolChest) QueryToolChest(org.apache.druid.query.QueryToolChest) Result(org.apache.druid.query.Result) SegmentDescriptor(org.apache.druid.query.SegmentDescriptor) TopNResultValue(org.apache.druid.query.topn.TopNResultValue) IncrementalIndex(org.apache.druid.segment.incremental.IncrementalIndex) OnheapIncrementalIndex(org.apache.druid.segment.incremental.OnheapIncrementalIndex) SegmentId(org.apache.druid.timeline.SegmentId) QueryRunner(org.apache.druid.query.QueryRunner) FinalizeResultsQueryRunner(org.apache.druid.query.FinalizeResultsQueryRunner) PerSegmentOptimizingQueryRunner(org.apache.druid.query.PerSegmentOptimizingQueryRunner) PerSegmentQueryOptimizationContext(org.apache.druid.query.PerSegmentQueryOptimizationContext) FinalizeResultsQueryRunner(org.apache.druid.query.FinalizeResultsQueryRunner) QueryableIndex(org.apache.druid.segment.QueryableIndex) DataGenerator(org.apache.druid.segment.generator.DataGenerator) InputRow(org.apache.druid.data.input.InputRow) File(java.io.File) Setup(org.openjdk.jmh.annotations.Setup)

Example 39 with QueryableIndex

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

the class DoublesSketchSqlAggregatorTest method createQuerySegmentWalker.

@Override
public SpecificSegmentsQuerySegmentWalker createQuerySegmentWalker() throws IOException {
    DoublesSketchModule.registerSerde();
    final QueryableIndex index = IndexBuilder.create(CalciteTests.getJsonMapper()).tmpDir(temporaryFolder.newFolder()).segmentWriteOutMediumFactory(OffHeapMemorySegmentWriteOutMediumFactory.instance()).schema(new IncrementalIndexSchema.Builder().withMetrics(new CountAggregatorFactory("cnt"), new DoubleSumAggregatorFactory("m1", "m1"), new DoublesSketchAggregatorFactory("qsketch_m1", "m1", 128)).withRollup(false).build()).rows(CalciteTests.ROWS1).buildMMappedIndex();
    return new SpecificSegmentsQuerySegmentWalker(conglomerate).add(DataSegment.builder().dataSource(CalciteTests.DATASOURCE1).interval(index.getDataInterval()).version("1").shardSpec(new LinearShardSpec(0)).size(0).build(), index);
}
Also used : CountAggregatorFactory(org.apache.druid.query.aggregation.CountAggregatorFactory) DoubleSumAggregatorFactory(org.apache.druid.query.aggregation.DoubleSumAggregatorFactory) SpecificSegmentsQuerySegmentWalker(org.apache.druid.sql.calcite.util.SpecificSegmentsQuerySegmentWalker) QueryableIndex(org.apache.druid.segment.QueryableIndex) LinearShardSpec(org.apache.druid.timeline.partition.LinearShardSpec) IndexBuilder(org.apache.druid.segment.IndexBuilder) DoublesSketchAggregatorFactory(org.apache.druid.query.aggregation.datasketches.quantiles.DoublesSketchAggregatorFactory)

Example 40 with QueryableIndex

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

the class HllSketchSqlAggregatorTest method createQuerySegmentWalker.

@Override
public SpecificSegmentsQuerySegmentWalker createQuerySegmentWalker() throws IOException {
    HllSketchModule.registerSerde();
    final QueryableIndex index = IndexBuilder.create().tmpDir(temporaryFolder.newFolder()).segmentWriteOutMediumFactory(OffHeapMemorySegmentWriteOutMediumFactory.instance()).schema(new IncrementalIndexSchema.Builder().withMetrics(new CountAggregatorFactory("cnt"), new DoubleSumAggregatorFactory("m1", "m1"), new HllSketchBuildAggregatorFactory("hllsketch_dim1", "dim1", null, null, ROUND)).withRollup(false).build()).rows(CalciteTests.ROWS1).buildMMappedIndex();
    walker = new SpecificSegmentsQuerySegmentWalker(conglomerate).add(DataSegment.builder().dataSource(CalciteTests.DATASOURCE1).interval(index.getDataInterval()).version("1").shardSpec(new LinearShardSpec(0)).size(0).build(), index);
    return walker;
}
Also used : CountAggregatorFactory(org.apache.druid.query.aggregation.CountAggregatorFactory) DoubleSumAggregatorFactory(org.apache.druid.query.aggregation.DoubleSumAggregatorFactory) SpecificSegmentsQuerySegmentWalker(org.apache.druid.sql.calcite.util.SpecificSegmentsQuerySegmentWalker) QueryableIndex(org.apache.druid.segment.QueryableIndex) LinearShardSpec(org.apache.druid.timeline.partition.LinearShardSpec) IndexBuilder(org.apache.druid.segment.IndexBuilder) HllSketchBuildAggregatorFactory(org.apache.druid.query.aggregation.datasketches.hll.HllSketchBuildAggregatorFactory)

Aggregations

QueryableIndex (org.apache.druid.segment.QueryableIndex)67 File (java.io.File)29 ArrayList (java.util.ArrayList)29 IncrementalIndex (org.apache.druid.segment.incremental.IncrementalIndex)18 DataSegment (org.apache.druid.timeline.DataSegment)18 QueryableIndexSegment (org.apache.druid.segment.QueryableIndexSegment)17 IOException (java.io.IOException)15 LinearShardSpec (org.apache.druid.timeline.partition.LinearShardSpec)15 IndexSpec (org.apache.druid.segment.IndexSpec)14 InputRow (org.apache.druid.data.input.InputRow)13 OnheapIncrementalIndex (org.apache.druid.segment.incremental.OnheapIncrementalIndex)13 CountAggregatorFactory (org.apache.druid.query.aggregation.CountAggregatorFactory)12 SpecificSegmentsQuerySegmentWalker (org.apache.druid.sql.calcite.util.SpecificSegmentsQuerySegmentWalker)12 List (java.util.List)10 Before (org.junit.Before)10 QueryRunner (org.apache.druid.query.QueryRunner)9 DoubleSumAggregatorFactory (org.apache.druid.query.aggregation.DoubleSumAggregatorFactory)9 IndexBuilder (org.apache.druid.segment.IndexBuilder)9 Nullable (javax.annotation.Nullable)8 ISE (org.apache.druid.java.util.common.ISE)8