Examples with GeneratorSchemaInfo org.apache.druid.segment.generator.GeneratorSchemaInfo used on opensource projects

Example 6 with GeneratorSchemaInfo

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

the class SqlBenchmark method setup.

@Setup(Level.Trial)
public void setup() {
    final GeneratorSchemaInfo schemaInfo = GeneratorBasicSchemas.SCHEMA_MAP.get("basic");
    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);
    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), createOperatorTable(), CalciteTests.createExprMacroTable(), plannerConfig, AuthTestUtils.TEST_AUTHORIZER_MAPPER, CalciteTests.getJsonMapper(), CalciteTests.DRUID_SCHEMA_NAME);
}

Example 7 with GeneratorSchemaInfo

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

the class SqlVsNativeBenchmark method setup.

@Setup(Level.Trial)
public void setup() {
    this.closer = Closer.create();
    final GeneratorSchemaInfo schemaInfo = GeneratorBasicSchemas.SCHEMA_MAP.get("basic");
    final DataSegment dataSegment = DataSegment.builder().dataSource("foo").interval(schemaInfo.getDataInterval()).version("1").shardSpec(new LinearShardSpec(0)).size(0).build();
    final SegmentGenerator segmentGenerator = closer.register(new SegmentGenerator());
    log.info("Starting benchmark setup using tmpDir[%s], rows[%,d].", segmentGenerator.getCacheDir(), rowsPerSegment);
    final QueryableIndex index = segmentGenerator.generate(dataSegment, schemaInfo, Granularities.NONE, rowsPerSegment);
    final QueryRunnerFactoryConglomerate conglomerate = QueryStackTests.createQueryRunnerFactoryConglomerate(closer);
    final PlannerConfig plannerConfig = new PlannerConfig();
    this.walker = closer.register(new SpecificSegmentsQuerySegmentWalker(conglomerate).add(dataSegment, index));
    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);
    groupByQuery = GroupByQuery.builder().setDataSource("foo").setInterval(Intervals.ETERNITY).setDimensions(new DefaultDimensionSpec("dimZipf", "d0"), new DefaultDimensionSpec("dimSequential", "d1")).setAggregatorSpecs(new CountAggregatorFactory("c")).setGranularity(Granularities.ALL).build();
    sqlQuery = "SELECT\n" + "  dimZipf AS d0," + "  dimSequential AS d1,\n" + "  COUNT(*) AS c\n" + "FROM druid.foo\n" + "GROUP BY dimZipf, dimSequential";
}

Example 8 with GeneratorSchemaInfo

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

the class IndexedTableJoinCursorBenchmark method makeQueryableIndexSegment.

public static QueryableIndexSegment makeQueryableIndexSegment(Closer closer, String dataSource, int rowsPerSegment) {
    final List<GeneratorColumnSchema> schemaColumnsInfo = ImmutableList.of(GeneratorColumnSchema.makeSequential("stringKey", ValueType.STRING, false, 1, null, 0, rowsPerSegment), GeneratorColumnSchema.makeSequential("longKey", ValueType.LONG, false, 1, null, 0, rowsPerSegment), GeneratorColumnSchema.makeLazyZipf("string1", ValueType.STRING, false, 1, 0.1, 0, rowsPerSegment, 2.0), GeneratorColumnSchema.makeLazyZipf("string2", ValueType.STRING, false, 1, 0.3, 0, 1000000, 1.5), GeneratorColumnSchema.makeLazyZipf("string3", ValueType.STRING, false, 1, 0.12, 0, 1000, 1.25), GeneratorColumnSchema.makeLazyZipf("string4", ValueType.STRING, false, 1, 0.22, 0, 12000, 3.0), GeneratorColumnSchema.makeLazyZipf("string5", ValueType.STRING, false, 1, 0.05, 0, 33333, 1.8), GeneratorColumnSchema.makeLazyZipf("long1", ValueType.LONG, false, 1, 0.1, 0, 1001, 2.0), GeneratorColumnSchema.makeLazyZipf("long2", ValueType.LONG, false, 1, 0.01, 0, 666666, 2.2), GeneratorColumnSchema.makeLazyZipf("long3", ValueType.LONG, false, 1, 0.12, 0, 1000000, 2.5), GeneratorColumnSchema.makeLazyZipf("long4", ValueType.LONG, false, 1, 0.4, 0, 23, 1.2), GeneratorColumnSchema.makeLazyZipf("long5", ValueType.LONG, false, 1, 0.33, 0, 9999, 1.5), GeneratorColumnSchema.makeLazyZipf("double1", ValueType.DOUBLE, false, 1, 0.1, 0, 333, 2.2), GeneratorColumnSchema.makeLazyZipf("double2", ValueType.DOUBLE, false, 1, 0.01, 0, 4021, 2.5), GeneratorColumnSchema.makeLazyZipf("double3", ValueType.DOUBLE, false, 1, 0.41, 0, 90210, 4.0), GeneratorColumnSchema.makeLazyZipf("double4", ValueType.DOUBLE, false, 1, 0.5, 0, 5555555, 1.2), GeneratorColumnSchema.makeLazyZipf("double5", ValueType.DOUBLE, false, 1, 0.23, 0, 80, 1.8), GeneratorColumnSchema.makeLazyZipf("float1", ValueType.FLOAT, false, 1, 0.11, 0, 1000000, 1.7), GeneratorColumnSchema.makeLazyZipf("float2", ValueType.FLOAT, false, 1, 0.4, 0, 10, 1.5), GeneratorColumnSchema.makeLazyZipf("float3", ValueType.FLOAT, false, 1, 0.8, 0, 5000, 2.3), GeneratorColumnSchema.makeLazyZipf("float4", ValueType.FLOAT, false, 1, 0.999, 0, 14440, 2.0), GeneratorColumnSchema.makeLazyZipf("float5", ValueType.FLOAT, false, 1, 0.001, 0, 1029, 1.5));
    final List<AggregatorFactory> aggs = new ArrayList<>();
    aggs.add(new CountAggregatorFactory("rows"));
    final Interval interval = Intervals.of("2000-01-01/P1D");
    final GeneratorSchemaInfo schema = new GeneratorSchemaInfo(schemaColumnsInfo, aggs, interval, false);
    final DataSegment dataSegment = DataSegment.builder().dataSource(dataSource).interval(schema.getDataInterval()).version("1").shardSpec(new LinearShardSpec(0)).size(0).build();
    final QueryableIndex index = closer.register(new SegmentGenerator()).generate(dataSegment, schema, Granularities.NONE, rowsPerSegment);
    return closer.register(new QueryableIndexSegment(index, SegmentId.dummy(dataSource)));
}

Example 9 with GeneratorSchemaInfo

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

the class TopNTypeInterfaceBenchmark method setupQueries.

private void setupQueries() {
    // queries for the basic schema
    Map<String, TopNQueryBuilder> basicQueries = new LinkedHashMap<>();
    GeneratorSchemaInfo basicSchema = GeneratorBasicSchemas.SCHEMA_MAP.get("basic");
    {
        // basic.A
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(basicSchema.getDataInterval()));
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        queryAggs.add(new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential"));
        queryAggs.add(new LongMaxAggregatorFactory("maxLongUniform", "maxLongUniform"));
        queryAggs.add(new DoubleSumAggregatorFactory("sumFloatNormal", "sumFloatNormal"));
        queryAggs.add(new DoubleMinAggregatorFactory("minFloatZipf", "minFloatZipf"));
        queryAggs.add(new HyperUniquesAggregatorFactory("hyperUniquesMet", "hyper"));
        // Use an IdentityExtractionFn to force usage of HeapBasedTopNAlgorithm
        TopNQueryBuilder queryBuilderString = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension(new ExtractionDimensionSpec("dimSequential", "dimSequential", IdentityExtractionFn.getInstance())).metric("sumFloatNormal").intervals(intervalSpec).aggregators(queryAggs);
        // HeapBasedTopNAlgorithm is always used for numeric columns
        TopNQueryBuilder queryBuilderLong = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension("metLongUniform").metric("sumFloatNormal").intervals(intervalSpec).aggregators(queryAggs);
        TopNQueryBuilder queryBuilderFloat = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension("metFloatNormal").metric("sumFloatNormal").intervals(intervalSpec).aggregators(queryAggs);
        basicQueries.put("string", queryBuilderString);
        basicQueries.put("long", queryBuilderLong);
        basicQueries.put("float", queryBuilderFloat);
    }
    {
        // basic.numericSort
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(basicSchema.getDataInterval()));
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        queryAggs.add(new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential"));
        TopNQueryBuilder queryBuilderA = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension("dimUniform").metric(new DimensionTopNMetricSpec(null, StringComparators.NUMERIC)).intervals(intervalSpec).aggregators(queryAggs);
        basicQueries.put("numericSort", queryBuilderA);
    }
    {
        // basic.alphanumericSort
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(basicSchema.getDataInterval()));
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        queryAggs.add(new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential"));
        TopNQueryBuilder queryBuilderA = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension("dimUniform").metric(new DimensionTopNMetricSpec(null, StringComparators.ALPHANUMERIC)).intervals(intervalSpec).aggregators(queryAggs);
        basicQueries.put("alphanumericSort", queryBuilderA);
    }
    SCHEMA_QUERY_MAP.put("basic", basicQueries);
}

Example 10 with GeneratorSchemaInfo

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

the class ExpressionSelectorBenchmark method setup.

@Setup(Level.Trial)
public void setup() {
    this.closer = Closer.create();
    final GeneratorSchemaInfo schemaInfo = new GeneratorSchemaInfo(ImmutableList.of(GeneratorColumnSchema.makeZipf("n", ValueType.LONG, false, 1, 0d, 1000, 10000, 3d), GeneratorColumnSchema.makeZipf("s", ValueType.STRING, false, 1, 0d, 1000, 10000, 3d)), ImmutableList.of(), Intervals.of("2000/P1D"), false);
    final DataSegment dataSegment = DataSegment.builder().dataSource("foo").interval(schemaInfo.getDataInterval()).version("1").shardSpec(new LinearShardSpec(0)).size(0).build();
    final SegmentGenerator segmentGenerator = closer.register(new SegmentGenerator());
    this.index = closer.register(segmentGenerator.generate(dataSegment, schemaInfo, Granularities.HOUR, rowsPerSegment));
}