Examples with IndexScanPlanNode org.voltdb.plannodes.IndexScanPlanNode used on opensource projects

Example 1 with IndexScanPlanNode

use of org.voltdb.plannodes.IndexScanPlanNode in project voltdb by VoltDB.

the class ReplaceWithIndexCounter method recursivelyApply.

@Override
protected AbstractPlanNode recursivelyApply(AbstractPlanNode plan) {
    assert (plan != null);
    // depth first:
    //     find AggregatePlanNode with exactly one child
    //     where that child is an AbstractScanPlanNode.
    //     Replace any qualifying AggregatePlanNode / AbstractScanPlanNode pair
    //     with an IndexCountPlanNode or TableCountPlanNode
    ArrayList<AbstractPlanNode> children = new ArrayList<AbstractPlanNode>();
    for (int i = 0; i < plan.getChildCount(); i++) children.add(plan.getChild(i));
    for (AbstractPlanNode child : children) {
        // TODO this will break when children feed multiple parents
        AbstractPlanNode newChild = recursivelyApply(child);
        // Do a graft into the (parent) plan only if a replacement for a child was found.
        if (newChild == child) {
            continue;
        }
        boolean replaced = plan.replaceChild(child, newChild);
        assert (true == replaced);
    }
    // check for an aggregation of the right form
    if ((plan instanceof AggregatePlanNode) == false)
        return plan;
    assert (plan.getChildCount() == 1);
    AggregatePlanNode aggplan = (AggregatePlanNode) plan;
    // ENG-6131 fixed here.
    if (!(aggplan.isTableCountStar() || aggplan.isTableNonDistinctCountConstant() || aggplan.isTableCountNonDistinctNullableColumn())) {
        return plan;
    }
    AbstractPlanNode child = plan.getChild(0);
    // A table count can replace a seq scan only if it has no predicates.
    if (child instanceof SeqScanPlanNode) {
        if (((SeqScanPlanNode) child).getPredicate() != null) {
            return plan;
        }
        AbstractExpression postPredicate = aggplan.getPostPredicate();
        if (postPredicate != null) {
            List<AbstractExpression> aggList = postPredicate.findAllAggregateSubexpressions();
            boolean allCountStar = true;
            for (AbstractExpression expr : aggList) {
                if (expr.getExpressionType() != ExpressionType.AGGREGATE_COUNT_STAR) {
                    allCountStar = false;
                    break;
                }
            }
            if (allCountStar) {
                return plan;
            }
        }
        if (hasInlineLimit(aggplan)) {
            // table count EE executor does not handle inline limit stuff
            return plan;
        }
        return new TableCountPlanNode((AbstractScanPlanNode) child, aggplan);
    }
    // Otherwise, optimized counts only replace particular cases of index scan.
    if ((child instanceof IndexScanPlanNode) == false)
        return plan;
    IndexScanPlanNode isp = (IndexScanPlanNode) child;
    // Guard against (possible future?) cases of indexable subquery.
    if (((IndexScanPlanNode) child).isSubQuery()) {
        return plan;
    }
    // except those (post-)predicates are artifact predicates we added for reverse scan purpose only
    if (isp.getPredicate() != null && !isp.isPredicatesOptimizableForAggregate()) {
        return plan;
    }
    // With no start or end keys, there's not much a counting index can do.
    if (isp.getEndExpression() == null && isp.getSearchKeyExpressions().size() == 0) {
        if (hasInlineLimit(aggplan)) {
            return plan;
        }
        return new TableCountPlanNode(isp, aggplan);
    }
    // check for the index's support for counting
    Index idx = isp.getCatalogIndex();
    if (!idx.getCountable()) {
        return plan;
    }
    // The core idea is that counting index needs to know the start key and end key to
    // jump to to get counts instead of actually doing any scanning.
    // Options to be determined are:
    // - whether each of the start/end keys is missing, partial (a prefix of a compund key), or complete,
    // - whether the count should include or exclude entries exactly matching each of the start/end keys.
    // Not all combinations of these options are supported;
    // unsupportable cases cause the factory method to return null.
    IndexCountPlanNode countingPlan = IndexCountPlanNode.createOrNull(isp, aggplan);
    if (countingPlan == null) {
        return plan;
    }
    return countingPlan;
}

Example 2 with IndexScanPlanNode

use of org.voltdb.plannodes.IndexScanPlanNode in project voltdb by VoltDB.

the class ReplaceWithIndexLimit method recursivelyApply.

@Override
protected AbstractPlanNode recursivelyApply(AbstractPlanNode plan) {
    assert (plan != null);
    // depth first:
    //     Find AggregatePlanNode with exactly one child
    //     where that child is an AbstractScanPlanNode.
    //     Replace qualifying SeqScanPlanNode with an
    //     IndexScanPlanNode with an inlined LimitPlanNode;
    //     or appending the LimitPlanNode to the existing
    //     qualified IndexScanPlanNode.
    ArrayList<AbstractPlanNode> children = new ArrayList<AbstractPlanNode>();
    for (int i = 0; i < plan.getChildCount(); i++) children.add(plan.getChild(i));
    for (AbstractPlanNode child : children) {
        // TODO this will break when children feed multiple parents
        AbstractPlanNode newChild = recursivelyApply(child);
        // Do a graft into the (parent) plan only if a replacement for a child was found.
        if (newChild == child) {
            continue;
        }
        child.removeFromGraph();
        plan.addAndLinkChild(newChild);
    }
    // check for an aggregation of the right form
    if ((plan instanceof AggregatePlanNode) == false)
        return plan;
    assert (plan.getChildCount() == 1);
    AggregatePlanNode aggplan = (AggregatePlanNode) plan;
    // handle one single min() / max() now
    // TODO: combination of [min(), max(), count()]
    SortDirectionType sortDirection = SortDirectionType.INVALID;
    if (aggplan.isTableMin()) {
        sortDirection = SortDirectionType.ASC;
    } else if (aggplan.isTableMax()) {
        sortDirection = SortDirectionType.DESC;
    } else {
        return plan;
    }
    AbstractPlanNode child = plan.getChild(0);
    AbstractExpression aggExpr = aggplan.getFirstAggregateExpression();
    // for a SEQSCAN, replace it with a INDEXSCAN node with an inline LIMIT plan node
    if (child instanceof SeqScanPlanNode) {
        // should have other index access plan if any qualified index found for the predicate
        if (((SeqScanPlanNode) child).getPredicate() != null) {
            return plan;
        }
        if (((AbstractScanPlanNode) child).isSubQuery()) {
            return plan;
        }
        // create an empty bindingExprs list, used for store (possible) bindings for adHoc query
        ArrayList<AbstractExpression> bindings = new ArrayList<AbstractExpression>();
        Index ret = findQualifiedIndex(((SeqScanPlanNode) child), aggExpr, bindings);
        if (ret == null) {
            return plan;
        } else {
            // 1. create one INDEXSCAN plan node with inlined LIMIT
            // and replace the SEQSCAN node with it
            // 2. we know which end row we want to fetch, so it's safe to
            // specify sorting direction here
            IndexScanPlanNode ispn = new IndexScanPlanNode((SeqScanPlanNode) child, aggplan, ret, sortDirection);
            ispn.setBindings(bindings);
            assert (ispn.getSearchKeyExpressions().size() == 0);
            if (sortDirection == SortDirectionType.ASC) {
                assert (aggplan.isTableMin());
                ispn.setSkipNullPredicate(0);
            }
            LimitPlanNode lpn = new LimitPlanNode();
            lpn.setLimit(1);
            lpn.setOffset(0);
            ispn.addInlinePlanNode(lpn);
            // remove old SeqScan node and link the new generated IndexScan node
            plan.clearChildren();
            plan.addAndLinkChild(ispn);
            return plan;
        }
    }
    if ((child instanceof IndexScanPlanNode) == false) {
        return plan;
    }
    // already have the IndexScanPlanNode
    IndexScanPlanNode ispn = (IndexScanPlanNode) child;
    // we added for reverse scan purpose only
    if (((IndexScanPlanNode) child).getPredicate() != null && !((IndexScanPlanNode) child).isPredicatesOptimizableForAggregate()) {
        return plan;
    }
    // Guard against (possible future?) cases of indexable subquery.
    if (((AbstractScanPlanNode) child).isSubQuery()) {
        return plan;
    }
    // 2. Handle equality filters and one other comparison operator (<, <=, >, >=), see comments below
    if (ispn.getLookupType() != IndexLookupType.EQ && Math.abs(ispn.getSearchKeyExpressions().size() - ExpressionUtil.uncombinePredicate(ispn.getEndExpression()).size()) > 1) {
        return plan;
    }
    // exprs will be used as filterExprs to check the index
    // For forward scan, the initial value is endExprs and might be changed in different values in variant cases
    // For reverse scan, the initial value is initialExprs which is the "old" endExprs
    List<AbstractExpression> exprs;
    int numOfSearchKeys = ispn.getSearchKeyExpressions().size();
    if (ispn.getLookupType() == IndexLookupType.LT || ispn.getLookupType() == IndexLookupType.LTE) {
        exprs = ExpressionUtil.uncombinePredicate(ispn.getInitialExpression());
        numOfSearchKeys -= 1;
    } else {
        exprs = ExpressionUtil.uncombinePredicate(ispn.getEndExpression());
    }
    int numberOfExprs = exprs.size();
    /* Retrieve the index expressions from the target index. (ENG-8819, Ethan)
         * This is because we found that for the following two queries:
         *     #1: explain select max(c2/2) from t where c1=1 and c2/2<=3;
         *     #2: explain select max(c2/2) from t where c1=1 and c2/2<=?;
         * We can get an inline limit 1 for #2 but not for #1. This is because all constants in #1 got parameterized.
         * The result is that the query cannot pass the bindingToIndexedExpression() tests below
         * because we lost all the constant value expressions (cannot attempt to bind a pve to a pve!).
         * Those constant values expressions can only be accessed from the idnex.
         * We will not add those bindings to the ispn.getBindings() here because they will be added anyway in checkIndex().
         * PS: For this case (i.e. index on expressions), checkIndex() will call checkExpressionIndex(),
         * where bindings will be added.
         */
    Index indexToUse = ispn.getCatalogIndex();
    String tableAlias = ispn.getTargetTableAlias();
    List<AbstractExpression> indexedExprs = null;
    if (!indexToUse.getExpressionsjson().isEmpty()) {
        StmtTableScan tableScan = m_parsedStmt.getStmtTableScanByAlias(tableAlias);
        try {
            indexedExprs = AbstractExpression.fromJSONArrayString(indexToUse.getExpressionsjson(), tableScan);
        } catch (JSONException e) {
            e.printStackTrace();
            assert (false);
            return plan;
        }
    }
    /* If there is only 1 difference between searchkeyExprs and endExprs,
         * 1. trivial filters can be discarded, 2 possibilities:
         *      a. SELECT MIN(X) FROM T WHERE [other prefix filters] X < / <= ?
         *         <=> SELECT MIN(X) FROM T WHERE [other prefix filters] && the X < / <= ? filter
         *      b. SELECT MAX(X) FROM T WHERE X > / >= ?
         *         <=> SELECT MAX(X) FROM T with post-filter
         * 2. filter should act as equality filter, 2 possibilities
         *      SELECT MIN(X) FROM T WHERE [other prefix filters] X > / >= ?
         *      SELECT MAX(X) FROM T WHERE [other prefix filters] X < / <= ?

         * check if there is other filters for SELECT MAX(X) FROM T WHERE [other prefix filter AND ] X > / >= ?
         * but we should allow SELECT MAX(X) FROM T WHERE X = ?

         * This is for queries having MAX() but no ORDER BY. (ENG-8819, Ethan)
         * sortDirection == DESC if max, ASC if min. ispn.getSortDirection() == INVALID if no ORDER BY. */
    if (sortDirection == SortDirectionType.DESC && ispn.getSortDirection() == SortDirectionType.INVALID) {
        /* numberOfExprs = exprs.size(), exprs are initial expressions for reversed index scans (lookupType LT, LTE),
             * are end expressions for forward index scans (lookupType GT, GTE, EQ).
             * Note, lookupType doesn't decide the scan direction for sure. MIN(X) where X < ? is still a forward scan.
             * X < ? will be a post filter for the scan rather than an initial expression. */
        if (numberOfExprs == 1) {
            // e.g.: explain select max(c2/2) from t where c2/2<=3;
            // In this case, as long as the where condition (exprs.get(0)) matches the aggregation argument, continue.
            AbstractExpression exprToBind = indexedExprs == null ? exprs.get(0).getLeft() : indexedExprs.get(0);
            if (aggExpr.bindingToIndexedExpression(exprToBind) == null) {
                return plan;
            }
        } else if (numberOfExprs > 1) {
            // ENG-4016: Optimization for query SELECT MAX(X) FROM T WHERE [other prefix filters] X < / <= ?
            // Just keep trying, don't return early.
            boolean earlyReturn = true;
            for (int i = 0; i < numberOfExprs; ++i) {
                AbstractExpression expr = exprs.get(i);
                AbstractExpression indexedExpr = indexedExprs == null ? expr.getLeft() : indexedExprs.get(i);
                if (aggExpr.bindingToIndexedExpression(indexedExpr) != null && (expr.getExpressionType() == ExpressionType.COMPARE_LESSTHANOREQUALTO || expr.getExpressionType() == ExpressionType.COMPARE_LESSTHAN || expr.getExpressionType() == ExpressionType.COMPARE_EQUAL)) {
                    earlyReturn = false;
                    break;
                }
            }
            if (earlyReturn) {
                return plan;
            }
        }
    }
    // have an upper bound: # of endingExpr is more than # of searchExpr
    if (numberOfExprs > numOfSearchKeys) {
        AbstractExpression lastEndExpr = exprs.get(numberOfExprs - 1);
        // check last ending condition, see whether it is
        //      SELECT MIN(X) FROM T WHERE [other prefix filters] X < / <= ? or
        // other filters will be checked later
        AbstractExpression exprToBind = indexedExprs == null ? lastEndExpr.getLeft() : indexedExprs.get(numberOfExprs - 1);
        if ((lastEndExpr.getExpressionType() == ExpressionType.COMPARE_LESSTHAN || lastEndExpr.getExpressionType() == ExpressionType.COMPARE_LESSTHANOREQUALTO) && aggExpr.bindingToIndexedExpression(exprToBind) != null) {
            exprs.remove(lastEndExpr);
        }
    }
    // and we can take advantage of that
    if (checkIndex(ispn.getCatalogIndex(), aggExpr, exprs, ispn.getBindings(), tableAlias)) {
        // we know which end we want to fetch, set the sort direction
        ispn.setSortDirection(sortDirection);
        // for SELECT MIN(X) FROM T WHERE [prefix filters] = ?
        if (numberOfExprs == numOfSearchKeys && sortDirection == SortDirectionType.ASC) {
            if (ispn.getLookupType() == IndexLookupType.GTE) {
                assert (aggplan.isTableMin());
                ispn.setSkipNullPredicate(numOfSearchKeys);
            }
        }
        // reset the IndexLookupType, remove "added" searchKey, add back to endExpression, and clear "added" predicate
        if (sortDirection == SortDirectionType.ASC && (ispn.getLookupType() == IndexLookupType.LT || ispn.getLookupType() == IndexLookupType.LTE)) {
            ispn.setLookupType(IndexLookupType.GTE);
            ispn.removeLastSearchKey();
            ispn.addEndExpression(ExpressionUtil.uncombinePredicate(ispn.getInitialExpression()).get(numberOfExprs - 1));
            ispn.setSkipNullPredicate(numOfSearchKeys);
            ispn.resetPredicate();
        }
        // add an inline LIMIT plan node to this index scan plan node
        LimitPlanNode lpn = new LimitPlanNode();
        lpn.setLimit(1);
        lpn.setOffset(0);
        ispn.addInlinePlanNode(lpn);
        //                                              |__LimitPlanNode
        if (sortDirection == SortDirectionType.DESC && !ispn.getSearchKeyExpressions().isEmpty() && exprs.isEmpty() && ExpressionUtil.uncombinePredicate(ispn.getInitialExpression()).isEmpty()) {
            AbstractExpression newPredicate = new ComparisonExpression();
            if (ispn.getLookupType() == IndexLookupType.GT)
                newPredicate.setExpressionType(ExpressionType.COMPARE_GREATERTHAN);
            if (ispn.getLookupType() == IndexLookupType.GTE)
                newPredicate.setExpressionType(ExpressionType.COMPARE_GREATERTHANOREQUALTO);
            newPredicate.setRight(ispn.getSearchKeyExpressions().get(0));
            newPredicate.setLeft(aggExpr);
            newPredicate.setValueType(aggExpr.getValueType());
            ispn.clearSearchKeyExpression();
            aggplan.setPrePredicate(newPredicate);
        }
    }
    return plan;
}

Example 3 with IndexScanPlanNode

use of org.voltdb.plannodes.IndexScanPlanNode in project voltdb by VoltDB.

the class TestSelfJoins method testIndexedSelfJoin.

public void testIndexedSelfJoin() {
    AbstractPlanNode.enableVerboseExplainForDebugging();
    IndexScanPlanNode c;
    AbstractPlanNode apn;
    AbstractPlanNode pn;
    NestLoopIndexPlanNode nlij;
    List<AbstractExpression> searchKeys;
    // SELF JOIN using two different indexes on the same table
    // sometimes with a surviving sort ordering that supports GROUP BY and/or ORDER BY.
    apn = compile("select * FROM R2 A, R2 B WHERE A.A = B.A AND B.C > 1 ORDER BY B.C");
    //* for debug */ System.out.println(apn.toExplainPlanString());
    // Some day, the wasteful projection node will not be here to skip.
    pn = apn.getChild(0).getChild(0);
    assertTrue(pn instanceof NestLoopIndexPlanNode);
    nlij = (NestLoopIndexPlanNode) pn;
    assertNull(nlij.getPreJoinPredicate());
    assertNull(nlij.getJoinPredicate());
    assertNull(nlij.getWherePredicate());
    assertEquals(1, nlij.getChildCount());
    c = (IndexScanPlanNode) nlij.getChild(0);
    assertNull(c.getPredicate());
    assertEquals(IndexLookupType.GT, c.getLookupType());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof ConstantValueExpression);
    c = (IndexScanPlanNode) nlij.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    assertEquals(IndexLookupType.GTE, c.getLookupType());
    assertNull(c.getPredicate());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof TupleValueExpression);
    apn = compile("select * FROM R2 A, R2 B WHERE A.A = B.A AND B.C > 1 ORDER BY B.A, B.C");
    //* for debug */ System.out.println(apn.toExplainPlanString());
    // Some day, the wasteful projection node will not be here to skip.
    pn = apn.getChild(0).getChild(0);
    assertTrue(pn instanceof OrderByPlanNode);
    pn = pn.getChild(0);
    assertTrue(pn instanceof NestLoopIndexPlanNode);
    nlij = (NestLoopIndexPlanNode) pn;
    assertNull(nlij.getPreJoinPredicate());
    assertNull(nlij.getJoinPredicate());
    assertNull(nlij.getWherePredicate());
    assertEquals(1, nlij.getChildCount());
    c = (IndexScanPlanNode) nlij.getChild(0);
    assertNull(c.getPredicate());
    assertEquals(IndexLookupType.GT, c.getLookupType());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof ConstantValueExpression);
    c = (IndexScanPlanNode) nlij.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    assertEquals(IndexLookupType.GTE, c.getLookupType());
    assertNull(c.getPredicate());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof TupleValueExpression);
    apn = compile("select * FROM R2 A, R2 B WHERE A.A = B.A AND B.A > 1 ORDER BY B.A, B.C");
    //* for debug */ System.out.println(apn.toExplainPlanString());
    // Some day, the wasteful projection node will not be here to skip.
    pn = apn.getChild(0).getChild(0);
    assertTrue(pn instanceof NestLoopIndexPlanNode);
    nlij = (NestLoopIndexPlanNode) pn;
    assertNull(nlij.getPreJoinPredicate());
    assertNull(nlij.getJoinPredicate());
    assertNull(nlij.getWherePredicate());
    assertEquals(1, nlij.getChildCount());
    assertTrue(nlij.getChild(0) instanceof IndexScanPlanNode);
    c = (IndexScanPlanNode) nlij.getChild(0);
    assertEquals(IndexLookupType.GT, c.getLookupType());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof ConstantValueExpression);
    c = (IndexScanPlanNode) nlij.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    assertEquals(IndexLookupType.GTE, c.getLookupType());
    assertNull(c.getPredicate());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof TupleValueExpression);
    apn = compile("select B.C, MAX(A.C) FROM R2 A, R2 B WHERE A.A = B.A AND B.C > 1 GROUP BY B.C ORDER BY B.C");
    //* for debug */ System.out.println(apn.toExplainPlanString());
    // Some day, the wasteful projection node will not be here to skip.
    pn = apn.getChild(0);
    assertNotNull(AggregatePlanNode.getInlineAggregationNode(pn));
    assertTrue(pn instanceof NestLoopIndexPlanNode);
    nlij = (NestLoopIndexPlanNode) pn;
    assertNull(nlij.getPreJoinPredicate());
    assertNull(nlij.getJoinPredicate());
    assertNull(nlij.getWherePredicate());
    assertEquals(1, nlij.getChildCount());
    c = (IndexScanPlanNode) nlij.getChild(0);
    assertNull(c.getPredicate());
    assertEquals(IndexLookupType.GT, c.getLookupType());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof ConstantValueExpression);
    c = (IndexScanPlanNode) nlij.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    assertEquals(IndexLookupType.GTE, c.getLookupType());
    assertNull(c.getPredicate());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof TupleValueExpression);
    apn = compile("select B.C, B.A FROM R2 A, R2 B WHERE A.A = B.A AND B.C > 1 GROUP BY B.A, B.C ORDER BY B.A, B.C");
    //* for debug */ System.out.println(apn.toExplainPlanString());
    // Some day, the wasteful projection node will not be here to skip.
    pn = apn.getChild(0).getChild(0);
    assertTrue(pn instanceof OrderByPlanNode);
    pn = pn.getChild(0);
    assertNotNull(AggregatePlanNode.getInlineAggregationNode(pn));
    assertTrue(pn instanceof NestLoopIndexPlanNode);
    nlij = (NestLoopIndexPlanNode) pn;
    assertNull(nlij.getPreJoinPredicate());
    assertNull(nlij.getJoinPredicate());
    assertNull(nlij.getWherePredicate());
    assertEquals(1, nlij.getChildCount());
    c = (IndexScanPlanNode) nlij.getChild(0);
    assertNull(c.getPredicate());
    assertEquals(IndexLookupType.GT, c.getLookupType());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof ConstantValueExpression);
    c = (IndexScanPlanNode) nlij.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    assertEquals(IndexLookupType.GTE, c.getLookupType());
    assertNull(c.getPredicate());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof TupleValueExpression);
    apn = compile("select B.C, B.A FROM R2 A, R2 B WHERE A.A = B.A AND B.A > 1 GROUP BY B.A, B.C ORDER BY B.A, B.C");
    //* for debug */ System.out.println(apn.toExplainPlanString());
    pn = apn.getChild(0);
    assertNotNull(AggregatePlanNode.getInlineAggregationNode(pn));
    assertTrue(pn instanceof NestLoopIndexPlanNode);
    nlij = (NestLoopIndexPlanNode) pn;
    assertNull(nlij.getPreJoinPredicate());
    assertNull(nlij.getJoinPredicate());
    assertNull(nlij.getWherePredicate());
    assertEquals(1, nlij.getChildCount());
    assertTrue(nlij.getChild(0) instanceof IndexScanPlanNode);
    c = (IndexScanPlanNode) nlij.getChild(0);
    assertEquals(IndexLookupType.GT, c.getLookupType());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof ConstantValueExpression);
    c = (IndexScanPlanNode) nlij.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    assertEquals(IndexLookupType.GTE, c.getLookupType());
    assertNull(c.getPredicate());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof TupleValueExpression);
    // Here's a case that can't be optimized because it purposely uses the "wrong" column
    // in the GROUP BY and ORDER BY.
    apn = compile("select B.C FROM R2 A, R2 B WHERE B.A = A.A AND B.C > 1 GROUP BY A.A, B.C ORDER BY A.A, B.C");
    //* for debug */ System.out.println(apn.toExplainPlanString());
    // Complex ORDER BY case: GROUP BY column that is not in the display column list
    pn = apn.getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    pn = pn.getChild(0);
    assertTrue(pn instanceof OrderByPlanNode);
    pn = pn.getChild(0);
    assertNotNull(AggregatePlanNode.getInlineAggregationNode(pn));
    assertTrue(pn instanceof NestLoopIndexPlanNode);
    nlij = (NestLoopIndexPlanNode) pn;
    assertNull(nlij.getPreJoinPredicate());
    assertNull(nlij.getJoinPredicate());
    assertNull(nlij.getWherePredicate());
    assertEquals(1, nlij.getChildCount());
    c = (IndexScanPlanNode) nlij.getChild(0);
    assertNull(c.getPredicate());
    assertEquals(IndexLookupType.GT, c.getLookupType());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof ConstantValueExpression);
    c = (IndexScanPlanNode) nlij.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    assertEquals(IndexLookupType.GTE, c.getLookupType());
    assertNull(c.getPredicate());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof TupleValueExpression);
    // Here's a case that can't be optimized because it purposely uses the "wrong" column
    // in the GROUP BY and ORDER BY.
    apn = compile("select B.C FROM R2 A, R2 B WHERE B.A = A.A AND B.C > 1 GROUP BY A.A, B.C ORDER BY B.C");
    //* for debug */ System.out.println(apn.toExplainPlanString());
    // Project-first case: GROUP BY column that is not in the order by or the display column list
    pn = apn.getChild(0);
    assertTrue(pn instanceof OrderByPlanNode);
    pn = pn.getChild(0);
    //TODO: This represents a missed optimization.
    // The projection could have been inlined.
    assertTrue(pn instanceof ProjectionPlanNode);
    pn = pn.getChild(0);
    assertNotNull(AggregatePlanNode.getInlineAggregationNode(pn));
    assertTrue(pn instanceof NestLoopIndexPlanNode);
    nlij = (NestLoopIndexPlanNode) pn;
    assertNull(nlij.getPreJoinPredicate());
    assertNull(nlij.getJoinPredicate());
    assertNull(nlij.getWherePredicate());
    assertEquals(1, nlij.getChildCount());
    c = (IndexScanPlanNode) nlij.getChild(0);
    assertNull(c.getPredicate());
    assertEquals(IndexLookupType.GT, c.getLookupType());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof ConstantValueExpression);
    c = (IndexScanPlanNode) nlij.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    assertEquals(IndexLookupType.GTE, c.getLookupType());
    assertNull(c.getPredicate());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof TupleValueExpression);
    // This variant shows that the GROUP BY can be a permutation of the sort order
    // without messing up the optimization
    apn = compile("select B.C, B.A FROM R2 A, R2 B WHERE A.A = B.A AND B.A > 1 GROUP BY B.C, B.A ORDER BY B.A, B.C");
    //* for debug */ System.out.println(apn.toExplainPlanString());
    // Some day, the wasteful projection node will not be here to skip.
    pn = apn.getChild(0);
    assertNotNull(AggregatePlanNode.getInlineAggregationNode(pn));
    assertTrue(pn instanceof NestLoopIndexPlanNode);
    nlij = (NestLoopIndexPlanNode) pn;
    assertNull(nlij.getPreJoinPredicate());
    assertNull(nlij.getJoinPredicate());
    assertNull(nlij.getWherePredicate());
    assertEquals(1, nlij.getChildCount());
    assertTrue(nlij.getChild(0) instanceof IndexScanPlanNode);
    c = (IndexScanPlanNode) nlij.getChild(0);
    assertEquals(IndexLookupType.GT, c.getLookupType());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof ConstantValueExpression);
    c = (IndexScanPlanNode) nlij.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    assertEquals(IndexLookupType.GTE, c.getLookupType());
    assertNull(c.getPredicate());
    searchKeys = c.getSearchKeyExpressions();
    assertEquals(1, searchKeys.size());
    assertTrue(searchKeys.get(0) instanceof TupleValueExpression);
}

Example 4 with IndexScanPlanNode

use of org.voltdb.plannodes.IndexScanPlanNode in project voltdb by VoltDB.

the class TestPlansSubQueries method testJoins.

public void testJoins() {
    AbstractPlanNode pn;
    List<AbstractPlanNode> planNodes;
    AbstractPlanNode nlpn;
    String sql, sqlNoSimplification, equivalentSql;
    // Left Outer join
    sql = "SELECT R1.A, R1.C FROM R1 LEFT JOIN (SELECT A, C FROM R2) T1 ON T1.C = R1.C ";
    sqlNoSimplification = "SELECT R1.A, R1.C FROM R1 LEFT JOIN (SELECT A, C FROM R2 LIMIT 10) T1 ON T1.C = R1.C ";
    equivalentSql = "SELECT R1.A, R1.C FROM R1 LEFT JOIN R2 T1 ON T1.C = R1.C ";
    planNodes = compileToFragments(sqlNoSimplification);
    assertEquals(1, planNodes.size());
    pn = planNodes.get(0).getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    nlpn = pn.getChild(0);
    assertTrue(nlpn instanceof NestLoopPlanNode);
    assertEquals(JoinType.LEFT, ((NestLoopPlanNode) nlpn).getJoinType());
    pn = nlpn.getChild(0);
    checkSeqScan(pn, "R1", "A", "C");
    pn = nlpn.getChild(1);
    checkSeqScan(pn, "T1", "C");
    pn = pn.getChild(0);
    checkSeqScan(pn, "R2", "A", "C");
    checkSubquerySimplification(sql, equivalentSql);
    // Join with partitioned tables
    // Join on coordinator: LEFT OUTER JOIN, replicated table on left side
    sql = "SELECT R1.A, R1.C FROM R1 LEFT JOIN (SELECT A, C FROM P1) T1 ON T1.C = R1.C ";
    sqlNoSimplification = "SELECT R1.A, R1.C FROM R1 LEFT JOIN (SELECT DISTINCT A, C FROM P1) T1 ON T1.C = R1.C ";
    equivalentSql = "SELECT R1.A, R1.C FROM R1 LEFT JOIN P1 T1 ON T1.C = R1.C ";
    planNodes = compileToFragments(sqlNoSimplification);
    assertEquals(2, planNodes.size());
    pn = planNodes.get(0).getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    nlpn = pn.getChild(0);
    assertTrue(nlpn instanceof NestLoopPlanNode);
    assertEquals(JoinType.LEFT, ((NestLoopPlanNode) nlpn).getJoinType());
    pn = nlpn.getChild(0);
    checkSeqScan(pn, "R1", "A", "C");
    pn = nlpn.getChild(1);
    assertEquals(PlanNodeType.RECEIVE, pn.getPlanNodeType());
    pn = planNodes.get(1);
    assertTrue(pn instanceof SendPlanNode);
    pn = pn.getChild(0);
    checkSeqScan(pn, "T1", "C");
    checkSubquerySimplification(sql, equivalentSql);
    // Group by inside of the subquery
    // whether it contains group by or not does not matter, because we check it by whether inner side is partitioned or not
    planNodes = compileToFragments("SELECT R1.A, R1.C FROM R1 LEFT JOIN (SELECT A, count(*) C FROM P1 GROUP BY A) T1 ON T1.C = R1.C ");
    assertEquals(2, planNodes.size());
    pn = planNodes.get(0).getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    nlpn = pn.getChild(0);
    assertTrue(nlpn instanceof NestLoopPlanNode);
    assertEquals(JoinType.LEFT, ((NestLoopPlanNode) nlpn).getJoinType());
    pn = nlpn.getChild(0);
    checkSeqScan(pn, "R1", "A", "C");
    pn = nlpn.getChild(1);
    assertTrue(pn instanceof ReceivePlanNode);
    pn = planNodes.get(1);
    assertTrue(pn instanceof SendPlanNode);
    pn = pn.getChild(0);
    checkSeqScan(pn, "T1", "C");
    pn = pn.getChild(0);
    checkPrimaryKeyIndexScan(pn, "P1", "A", "C");
    assertNotNull(pn.getInlinePlanNode(PlanNodeType.PROJECTION));
    // Using index scan for group by only: use serial aggregate instead hash aggregate
    assertNotNull(pn.getInlinePlanNode(PlanNodeType.AGGREGATE));
    // LEFT partition table
    planNodes = compileToFragments("SELECT T1.CC FROM P1 LEFT JOIN (SELECT A, count(*) CC FROM P2 GROUP BY A) T1 ON T1.A = P1.A ");
    assertEquals(2, planNodes.size());
    pn = planNodes.get(0).getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    pn = pn.getChild(0);
    assertTrue(pn instanceof ReceivePlanNode);
    pn = planNodes.get(1);
    assertTrue(pn instanceof SendPlanNode);
    nlpn = pn.getChild(0);
    assertTrue(nlpn instanceof NestLoopPlanNode);
    assertEquals(JoinType.LEFT, ((NestLoopPlanNode) nlpn).getJoinType());
    pn = nlpn.getChild(0);
    checkPrimaryKeyIndexScan(pn, "P1");
    pn = nlpn.getChild(1);
    checkSeqScan(pn, "T1");
    pn = pn.getChild(0);
    checkPrimaryKeyIndexScan(pn, "P2");
    assertNotNull(pn.getInlinePlanNode(PlanNodeType.PROJECTION));
    // Using index scan for group by only: use serial aggregate instead hash aggregate
    assertNotNull(pn.getInlinePlanNode(PlanNodeType.AGGREGATE));
    // Right outer join
    planNodes = compileToFragments("SELECT R1.A, R1.C FROM R1 RIGHT JOIN (SELECT A, count(*) C FROM P1 GROUP BY A) T1 ON T1.C = R1.C ");
    assertEquals(2, planNodes.size());
    pn = planNodes.get(0).getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    pn = pn.getChild(0);
    assertTrue(pn instanceof ReceivePlanNode);
    pn = planNodes.get(1);
    assertTrue(pn instanceof SendPlanNode);
    nlpn = pn.getChild(0);
    assertTrue(nlpn instanceof NestLoopPlanNode);
    assertEquals(JoinType.LEFT, ((NestLoopPlanNode) nlpn).getJoinType());
    pn = nlpn.getChild(1);
    checkSeqScan(pn, "R1", "A", "C");
    pn = nlpn.getChild(0);
    checkSeqScan(pn, "T1", "C");
    pn = pn.getChild(0);
    checkPrimaryKeyIndexScan(pn, "P1", "A", "C");
    assertNotNull(pn.getInlinePlanNode(PlanNodeType.PROJECTION));
    // Using index scan for group by only: use serial aggregate instead hash aggregate
    assertNotNull(pn.getInlinePlanNode(PlanNodeType.AGGREGATE));
    // RIGHT partition table
    planNodes = compileToFragments("SELECT T1.CC FROM P1 RIGHT JOIN (SELECT A, count(*) CC FROM P2 GROUP BY A) T1 ON T1.A = P1.A ");
    assertEquals(2, planNodes.size());
    pn = planNodes.get(0).getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    pn = pn.getChild(0);
    assertTrue(pn instanceof ReceivePlanNode);
    pn = planNodes.get(1);
    assertTrue(pn instanceof SendPlanNode);
    nlpn = pn.getChild(0);
    assertTrue(nlpn instanceof NestLoopIndexPlanNode);
    assertEquals(JoinType.LEFT, ((NestLoopIndexPlanNode) nlpn).getJoinType());
    pn = nlpn.getInlinePlanNode(PlanNodeType.INDEXSCAN);
    checkPrimaryKeyIndexScan(pn, "P1");
    pn = nlpn.getChild(0);
    checkSeqScan(pn, "T1");
    pn = pn.getChild(0);
    checkPrimaryKeyIndexScan(pn, "P2");
    assertNotNull(pn.getInlinePlanNode(PlanNodeType.PROJECTION));
    // Using index scan for group by only: use serial aggregate instead hash aggregate
    assertNotNull(pn.getInlinePlanNode(PlanNodeType.AGGREGATE));
    // Join locally: inner join case for subselects
    sql = "SELECT R1.A, R1.C FROM R1 INNER JOIN (SELECT A, C FROM P1) T1 ON T1.C = R1.C ";
    sqlNoSimplification = "SELECT R1.A, R1.C FROM R1 INNER JOIN (SELECT DISTINCT A, C FROM P1) T1 ON T1.C = R1.C ";
    equivalentSql = "SELECT R1.A, R1.C FROM R1 INNER JOIN P1 T1 ON T1.C = R1.C ";
    planNodes = compileToFragments(sqlNoSimplification);
    assertEquals(2, planNodes.size());
    pn = planNodes.get(0).getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    pn = pn.getChild(0);
    assertTrue(pn instanceof ReceivePlanNode);
    pn = planNodes.get(1);
    assertTrue(pn instanceof SendPlanNode);
    nlpn = pn.getChild(0);
    assertTrue(nlpn instanceof NestLoopPlanNode);
    assertEquals(JoinType.INNER, ((NestLoopPlanNode) nlpn).getJoinType());
    pn = nlpn.getChild(0);
    checkSeqScan(pn, "R1", "A", "C");
    pn = nlpn.getChild(1);
    checkSeqScan(pn, "T1", "C");
    checkSubquerySimplification(sql, equivalentSql);
    // Two sub-queries. One is partitioned and the other one is replicated
    sql = "select A, AC FROM (SELECT A FROM R1) T1, (SELECT C AC FROM P1) T2 WHERE T1.A = T2.AC ";
    sqlNoSimplification = "select A, AC FROM (SELECT A FROM R1 LIMIT 10) T1, (SELECT DISTINCT A AC FROM P1) T2 WHERE T1.A = T2.AC ";
    equivalentSql = "select T1.A, T2.C AC FROM R1 T1, P1 T2 WHERE T1.A = T2.C ";
    planNodes = compileToFragments(sqlNoSimplification);
    assertEquals(2, planNodes.size());
    pn = planNodes.get(0).getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    pn = pn.getChild(0);
    assertTrue(pn instanceof ReceivePlanNode);
    pn = planNodes.get(1);
    assertTrue(pn instanceof SendPlanNode);
    nlpn = pn.getChild(0);
    assertTrue(nlpn instanceof NestLoopPlanNode);
    assertEquals(JoinType.INNER, ((NestLoopPlanNode) nlpn).getJoinType());
    pn = nlpn.getChild(0);
    checkSeqScan(pn, "T1", "A");
    pn = pn.getChild(0);
    checkSeqScan(pn, "R1", "A");
    pn = nlpn.getChild(1);
    checkSeqScan(pn, "T2", "AC");
    checkSubquerySimplification(sql, equivalentSql);
    // This is a single fragment plan because planner can detect "A = 3".
    // Join locally
    sql = "select A1, A2 FROM (SELECT A A1 FROM R1) T1, (SELECT A A2 FROM P1 where A = 3) T2 WHERE T1.A1 = T2.A2 ";
    sqlNoSimplification = "select A2, A1 FROM (SELECT DISTINCT A A1 FROM R1) T1, (SELECT DISTINCT A A2 FROM P1 where A = 3) T2 WHERE T1.A1 = T2.A2 ";
    equivalentSql = "select T1.A A1, T2.A A2 FROM R1 T1 join P1 T2 on T2.A = 3 and T1.A = T2.A";
    planNodes = compileToFragments(sqlNoSimplification);
    assertEquals(1, planNodes.size());
    pn = planNodes.get(0);
    assertTrue(pn instanceof SendPlanNode);
    pn = pn.getChild(0);
    assertTrue(pn instanceof ProjectionPlanNode);
    nlpn = pn.getChild(0);
    assertTrue(nlpn instanceof NestLoopPlanNode);
    pn = nlpn.getChild(0);
    checkSeqScan(pn, "T1", "A1");
    pn = nlpn.getChild(1);
    checkSeqScan(pn, "T2", "A2");
    pn = pn.getChild(0);
    checkPrimaryKeyIndexScan(pn, "P1", "A");
    assertEquals(2, ((IndexScanPlanNode) pn).getInlinePlanNodes().size());
    assertNotNull(((IndexScanPlanNode) pn).getInlinePlanNode(PlanNodeType.PROJECTION));
    assertNotNull(((IndexScanPlanNode) pn).getInlinePlanNode(PlanNodeType.AGGREGATE));
    checkSubquerySimplification(sql, equivalentSql);
    // More single partition detection
    planNodes = compileToFragments("select C FROM (SELECT P1.C FROM P1, P2 " + "WHERE P1.A = P2.A AND P1.A = 3) T1 ");
    assertEquals(1, planNodes.size());
    planNodes = compileToFragments("select T1.C FROM (SELECT P1.C FROM P1, P2 " + "WHERE P1.A = P2.A AND P1.A = 3) T1, R1 where T1.C > R1.C ");
    assertEquals(1, planNodes.size());
    planNodes = compileToFragments("select T1.C FROM (SELECT P1.C FROM P1, P2 " + "WHERE P1.A = P2.A AND P1.A = 3) T1, (select C FROM R1) T2 where T1.C > T2.C ");
    assertEquals(1, planNodes.size());
}

Example 5 with IndexScanPlanNode

use of org.voltdb.plannodes.IndexScanPlanNode in project voltdb by VoltDB.

the class TestPlansSubQueries method testExpressionSubqueryWithIndexScan.

/**
     * Expression subquery currently is not optimized to use any index. But this does not prevent the
     * parent query to use index for other purposes.
     */
public void testExpressionSubqueryWithIndexScan() {
    AbstractPlanNode pn;
    String sql;
    // INDEX on A, for sort order only
    sql = "SELECT A FROM R4 where A in (select A from R4 where A > 3) order by A;";
    pn = compile(sql);
    pn = pn.getChild(0);
    assertTrue(pn instanceof IndexScanPlanNode);
    assertEquals(0, ((IndexScanPlanNode) pn).getSearchKeyExpressions().size());
    assertNotNull(((IndexScanPlanNode) pn).getPredicate());
    // INDEX on A, uniquely match A = 4,
    sql = "SELECT A FROM R4 where A = 4 and C in (select A from R4 where A > 3);";
    pn = compile(sql);
    pn = pn.getChild(0);
    assertTrue(pn instanceof IndexScanPlanNode);
    assertEquals(1, ((IndexScanPlanNode) pn).getSearchKeyExpressions().size());
    AbstractExpression comp = ((IndexScanPlanNode) pn).getSearchKeyExpressions().get(0);
    assertEquals(ExpressionType.VALUE_CONSTANT, comp.getExpressionType());
    assertEquals("4", ((ConstantValueExpression) comp).getValue());
    assertNotNull(((IndexScanPlanNode) pn).getPredicate());
}