Examples with PARTITIONED com.facebook.presto.sql.planner.plan.JoinNode.DistributionType.PARTITIONED used on opensource projects

Example 1 with PARTITIONED

use of com.facebook.presto.sql.planner.plan.JoinNode.DistributionType.PARTITIONED in project presto by prestodb.

the class TestLogicalPlanner method testBroadcastCorrelatedSubqueryAvoidsRemoteExchangeBeforeAggregation.

@Test
public void testBroadcastCorrelatedSubqueryAvoidsRemoteExchangeBeforeAggregation() {
    Session broadcastJoin = Session.builder(this.getQueryRunner().getDefaultSession()).setSystemProperty(JOIN_DISTRIBUTION_TYPE, JoinDistributionType.BROADCAST.name()).setSystemProperty(FORCE_SINGLE_NODE_OUTPUT, Boolean.toString(false)).build();
    // make sure there is a remote exchange on the build side
    PlanMatchPattern joinBuildSideWithRemoteExchange = anyTree(node(JoinNode.class, anyTree(node(TableScanNode.class)), anyTree(exchange(REMOTE_STREAMING, ExchangeNode.Type.REPLICATE, anyTree(node(TableScanNode.class))))));
    // validates that there exists only one remote exchange
    Consumer<Plan> validateSingleRemoteExchange = plan -> assertEquals(countOfMatchingNodes(plan, node -> node instanceof ExchangeNode && ((ExchangeNode) node).getScope().isRemote()), 1);
    Consumer<Plan> validateSingleStreamingAggregation = plan -> assertEquals(countOfMatchingNodes(plan, node -> node instanceof AggregationNode && ((AggregationNode) node).getGroupingKeys().contains(new VariableReferenceExpression(Optional.empty(), "unique", BIGINT)) && ((AggregationNode) node).isStreamable()), 1);
    // region is unpartitioned, AssignUniqueId should provide satisfying partitioning for count(*) after LEFT JOIN
    assertPlanWithSession("SELECT (SELECT COUNT(*) FROM region r2 WHERE r2.regionkey > r1.regionkey) FROM region r1", broadcastJoin, false, joinBuildSideWithRemoteExchange, validateSingleRemoteExchange.andThen(validateSingleStreamingAggregation));
    // orders is naturally partitioned, AssignUniqueId should not overwrite its natural partitioning
    assertPlanWithSession("SELECT COUNT(COUNT) " + "FROM (SELECT o1.orderkey orderkey, (SELECT COUNT(*) FROM orders o2 WHERE o2.orderkey > o1.orderkey) COUNT FROM orders o1) " + "GROUP BY orderkey", broadcastJoin, false, joinBuildSideWithRemoteExchange, validateSingleRemoteExchange.andThen(validateSingleStreamingAggregation));
}

Example 2 with PARTITIONED

use of com.facebook.presto.sql.planner.plan.JoinNode.DistributionType.PARTITIONED in project presto by prestodb.

the class TestPredicatePushdown method testPredicatePushDownCreatesValidJoin.

@Test
public void testPredicatePushDownCreatesValidJoin() {
    RuleTester tester = new RuleTester();
    tester.assertThat(new PredicatePushDown(tester.getMetadata(), tester.getSqlParser())).on(p -> p.join(INNER, p.filter(p.comparison(OperatorType.EQUAL, p.variable("a1"), constant(1L, INTEGER)), p.values(p.variable("a1"))), p.values(p.variable("b1")), ImmutableList.of(new EquiJoinClause(p.variable("a1"), p.variable("b1"))), ImmutableList.of(), Optional.empty(), Optional.empty(), Optional.empty(), Optional.of(PARTITIONED), ImmutableMap.of())).matches(project(join(INNER, ImmutableList.of(), Optional.empty(), Optional.of(REPLICATED), project(filter("a1=1", values("a1"))), project(filter("1=b1", values("b1"))))));
}

Example 3 with PARTITIONED

use of com.facebook.presto.sql.planner.plan.JoinNode.DistributionType.PARTITIONED in project presto by prestodb.

the class TestDetermineJoinDistributionType method testPartitionFullOuterJoin.

@Test
public void testPartitionFullOuterJoin() {
    int aRows = 10_000;
    int bRows = 10;
    assertDetermineJoinDistributionType().setSystemProperty(JOIN_DISTRIBUTION_TYPE, JoinDistributionType.AUTOMATIC.name()).overrideStats("valuesA", PlanNodeStatsEstimate.builder().setOutputRowCount(aRows).addVariableStatistics(ImmutableMap.of(new VariableReferenceExpression(Optional.empty(), "A1", BIGINT), new VariableStatsEstimate(0, 100, 0, 640000, 100))).build()).overrideStats("valuesB", PlanNodeStatsEstimate.builder().setOutputRowCount(bRows).addVariableStatistics(ImmutableMap.of(new VariableReferenceExpression(Optional.empty(), "B1", BIGINT), new VariableStatsEstimate(0, 100, 0, 640000, 100))).build()).on(p -> p.join(FULL, p.values(new PlanNodeId("valuesA"), aRows, p.variable("A1", BIGINT)), p.values(new PlanNodeId("valuesB"), bRows, p.variable("B1", BIGINT)), ImmutableList.of(new JoinNode.EquiJoinClause(p.variable("A1", BIGINT), p.variable("B1", BIGINT))), ImmutableList.of(p.variable("A1", BIGINT), p.variable("B1", BIGINT)), Optional.empty())).matches(join(FULL, ImmutableList.of(equiJoinClause("A1", "B1")), Optional.empty(), Optional.of(PARTITIONED), values(ImmutableMap.of("A1", 0)), values(ImmutableMap.of("B1", 0))));
}

Example 4 with PARTITIONED

use of com.facebook.presto.sql.planner.plan.JoinNode.DistributionType.PARTITIONED in project presto by prestodb.

the class TestDetermineJoinDistributionType method testChoosesRightWhenFallsBackToSyntactic.

@Test
public void testChoosesRightWhenFallsBackToSyntactic() {
    int aRows = 1_000__00;
    int bRows = 1_0;
    assertDetermineJoinDistributionType(new CostComparator(75, 10, 15)).setSystemProperty(JOIN_DISTRIBUTION_TYPE, JoinDistributionType.AUTOMATIC.name()).setSystemProperty(JOIN_MAX_BROADCAST_TABLE_SIZE, "100MB").overrideStats("valuesA", PlanNodeStatsEstimate.builder().setOutputRowCount(aRows).addVariableStatistics(ImmutableMap.of(new VariableReferenceExpression(Optional.empty(), "A1", BIGINT), new VariableStatsEstimate(0, 100, 0, 640000, 100))).build()).overrideStats("valuesB", PlanNodeStatsEstimate.builder().setOutputRowCount(bRows).addVariableStatistics(ImmutableMap.of(new VariableReferenceExpression(Optional.empty(), "B1", BIGINT), new VariableStatsEstimate(0, 100, 0, 640000, 100))).build()).on(p -> p.join(RIGHT, p.values(new PlanNodeId("valuesA"), aRows, p.variable("A1", BIGINT)), p.values(new PlanNodeId("valuesB"), bRows, p.variable("B1", BIGINT)), ImmutableList.of(), ImmutableList.of(p.variable("A1", BIGINT), p.variable("B1", BIGINT)), Optional.empty())).matches(join(RIGHT, ImmutableList.of(), Optional.empty(), Optional.of(PARTITIONED), values(ImmutableMap.of("B1", 0)), values(ImmutableMap.of("A1", 0))));
}

Example 5 with PARTITIONED

use of com.facebook.presto.sql.planner.plan.JoinNode.DistributionType.PARTITIONED in project presto by prestodb.

the class TestDetermineJoinDistributionType method testReplicatesWhenNotRestricted.

@Test
public void testReplicatesWhenNotRestricted() {
    int aRows = 10_000;
    int bRows = 10;
    PlanNodeStatsEstimate probeSideStatsEstimate = PlanNodeStatsEstimate.builder().setOutputRowCount(aRows).addVariableStatistics(ImmutableMap.of(new VariableReferenceExpression(Optional.empty(), "A1", BIGINT), new VariableStatsEstimate(0, 100, 0, 640000, 10))).build();
    PlanNodeStatsEstimate buildSideStatsEstimate = PlanNodeStatsEstimate.builder().setOutputRowCount(bRows).addVariableStatistics(ImmutableMap.of(new VariableReferenceExpression(Optional.empty(), "B1", BIGINT), new VariableStatsEstimate(0, 100, 0, 640000, 10))).build();
    // B table is small enough to be replicated in AUTOMATIC_RESTRICTED mode
    assertDetermineJoinDistributionType().setSystemProperty(JOIN_DISTRIBUTION_TYPE, JoinDistributionType.AUTOMATIC.name()).setSystemProperty(JOIN_MAX_BROADCAST_TABLE_SIZE, "100MB").overrideStats("valuesA", probeSideStatsEstimate).overrideStats("valuesB", buildSideStatsEstimate).on(p -> p.join(INNER, p.values(new PlanNodeId("valuesA"), aRows, p.variable("A1", BIGINT)), p.values(new PlanNodeId("valuesB"), bRows, p.variable("B1", BIGINT)), ImmutableList.of(new JoinNode.EquiJoinClause(p.variable("A1", BIGINT), p.variable("B1", BIGINT))), ImmutableList.of(p.variable("A1", BIGINT), p.variable("B1", BIGINT)), Optional.empty())).matches(join(INNER, ImmutableList.of(equiJoinClause("A1", "B1")), Optional.empty(), Optional.of(REPLICATED), values(ImmutableMap.of("A1", 0)), values(ImmutableMap.of("B1", 0))));
    probeSideStatsEstimate = PlanNodeStatsEstimate.builder().setOutputRowCount(aRows).addVariableStatistics(ImmutableMap.of(new VariableReferenceExpression(Optional.empty(), "A1", BIGINT), new VariableStatsEstimate(0, 100, 0, 640000d * 10000, 10))).build();
    buildSideStatsEstimate = PlanNodeStatsEstimate.builder().setOutputRowCount(bRows).addVariableStatistics(ImmutableMap.of(new VariableReferenceExpression(Optional.empty(), "B1", BIGINT), new VariableStatsEstimate(0, 100, 0, 640000d * 10000, 10))).build();
    // B table exceeds AUTOMATIC_RESTRICTED limit therefore it is partitioned
    assertDetermineJoinDistributionType().setSystemProperty(JOIN_DISTRIBUTION_TYPE, JoinDistributionType.AUTOMATIC.name()).setSystemProperty(JOIN_MAX_BROADCAST_TABLE_SIZE, "100MB").overrideStats("valuesA", probeSideStatsEstimate).overrideStats("valuesB", buildSideStatsEstimate).on(p -> p.join(INNER, p.values(new PlanNodeId("valuesA"), aRows, p.variable("A1", BIGINT)), p.values(new PlanNodeId("valuesB"), bRows, p.variable("B1", BIGINT)), ImmutableList.of(new JoinNode.EquiJoinClause(p.variable("A1", BIGINT), p.variable("B1", BIGINT))), ImmutableList.of(p.variable("A1", BIGINT), p.variable("B1", BIGINT)), Optional.empty())).matches(join(INNER, ImmutableList.of(equiJoinClause("A1", "B1")), Optional.empty(), Optional.of(PARTITIONED), values(ImmutableMap.of("A1", 0)), values(ImmutableMap.of("B1", 0))));
}