1
0
Fork 0

Revert "Fixed applying late materialization with calc node between sort an limit (#10439)" (#10446)

This reverts commit e765137c3e.
This commit is contained in:
Andrey Abramov 2019-11-15 12:56:05 +03:00 committed by GitHub
parent e765137c3e
commit c6fc77858f
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
6 changed files with 17 additions and 75 deletions

View File

@ -293,7 +293,7 @@ void lateDocumentMaterializationArangoSearchRule(arangodb::aql::Optimizer* opt,
if (arangodb::aql::ExecutionNode::ENUMERATE_IRESEARCH_VIEW == loop->getType()) {
auto & viewNode = *EN::castTo<IResearchViewNode*>(loop);
if (viewNode.isLateMaterialized()) {
continue; // loop is already optimized
continue; //loop is aleady optimized
}
ExecutionNode* current = limitNode->getFirstDependency();
ExecutionNode* sortNode = nullptr;
@ -301,7 +301,6 @@ void lateDocumentMaterializationArangoSearchRule(arangodb::aql::Optimizer* opt,
// without document body usage before that node.
// this node could be appended with materializer
bool stopSearch = false;
bool stickToSortNode = false;
while (current != loop) {
switch (current->getType()) {
case arangodb::aql::ExecutionNode::SORT:
@ -322,21 +321,15 @@ void lateDocumentMaterializationArangoSearchRule(arangodb::aql::Optimizer* opt,
default: // make clang happy
break;
}
if (!stopSearch) {
if (sortNode != nullptr) {
::arangodb::containers::HashSet<Variable const*> currentUsedVars;
current->getVariablesUsedHere(currentUsedVars);
if (currentUsedVars.find(&viewNode.outVariable()) != currentUsedVars.end()) {
if (sortNode != nullptr) {
// we have a doc body used before selected SortNode. Forget it, let`s look for better sort to use
sortNode = nullptr;
// this limit node affects only closest sort, if this sort is invalid
// we need to check other limit node
stopSearch = true;
} else {
// we are between limit and sort nodes.
// late materialization could still be applied but we must insert MATERIALIZE node after sort not after limit
stickToSortNode = true;
}
// we have a doc body used before selected SortNode. Forget it, let`s look for better sort to use
sortNode = nullptr;
// this limit node affects only closest sort, if this sort is invalid
// we need to check other limit node
stopSearch = true;
}
}
if (stopSearch) {
@ -359,7 +352,7 @@ void lateDocumentMaterializationArangoSearchRule(arangodb::aql::Optimizer* opt,
// on cluster we need to materialize node stay close to sort node on db server (to avoid network hop for materialization calls)
// however on single server we move it to limit node to make materialization as lazy as possible
auto materializeDependency = ServerState::instance()->isCoordinator() || stickToSortNode ? sortNode : limitNode;
auto materializeDependency = ServerState::instance()->isCoordinator() ? sortNode : limitNode;
auto* dependencyParent = materializeDependency->getFirstParent();
TRI_ASSERT(dependencyParent);
dependencyParent->replaceDependency(materializeDependency, materializeNode);

View File

@ -36,7 +36,7 @@ class ExecutionPlan;
namespace iresearch {
/// @brief moves document materialization from view nodes to materialize nodes
/// @brief moves document materialization from view nodes to sort nodes
void lateDocumentMaterializationArangoSearchRule(arangodb::aql::Optimizer* opt,
std::unique_ptr<arangodb::aql::ExecutionPlan> plan,
arangodb::aql::OptimizerRule const& rule);

View File

@ -201,7 +201,7 @@ void arangodb::aql::lateDocumentMaterializationRule(arangodb::aql::Optimizer* op
if (arangodb::aql::ExecutionNode::INDEX == loop->getType()) {
auto indexNode = EN::castTo<IndexNode*>(loop);
if (indexNode->isLateMaterialized()) {
continue; // loop is already optimized
continue; // loop is aleady optimized
}
auto current = limitNode->getFirstDependency();
ExecutionNode* sortNode = nullptr;
@ -209,7 +209,6 @@ void arangodb::aql::lateDocumentMaterializationRule(arangodb::aql::Optimizer* op
// without document body usage before that node.
// this node could be appended with materializer
bool stopSearch = false;
bool stickToSortNode = false;
std::vector<NodeWithAttrs> nodesToChange;
TRI_idx_iid_t commonIndexId = 0; // use one index only
while (current != loop) {
@ -231,17 +230,9 @@ void arangodb::aql::lateDocumentMaterializationRule(arangodb::aql::Optimizer* op
} else if (!node.attrs.empty()) {
if (!attributesMatch(commonIndexId, indexNode, node)) {
// the node uses attributes which is not in index
if (nullptr == sortNode) {
// we are between limit and sort nodes.
// late materialization could still be applied but we must insert MATERIALIZE node after sort not after limit
stickToSortNode = true;
} else {
// this limit node affects only closest sort, if this sort is invalid
// we need to check other limit node
stopSearch = true;
}
stopSearch = true;
} else {
nodesToChange.emplace_back(std::move(node));
nodesToChange.emplace_back(node);
}
}
break;
@ -255,19 +246,15 @@ void arangodb::aql::lateDocumentMaterializationRule(arangodb::aql::Optimizer* op
default: // make clang happy
break;
}
#ifdef ARANGODB_ENABLE_MAINTAINER_MODE
// Currently only calculation nodes expected to use loop variable and we successfully replaced
// all references to loop variable. However if some other node types will begin to use
// loop variable assertion below will be triggered and this rule should be updated
if (!stopSearch && current->getType() != arangodb::aql::ExecutionNode::CALCULATION) {
if (sortNode != nullptr && current->getType() != arangodb::aql::ExecutionNode::CALCULATION) {
::arangodb::containers::HashSet<Variable const*> currentUsedVars;
current->getVariablesUsedHere(currentUsedVars);
if (currentUsedVars.find(indexNode->outVariable()) != currentUsedVars.end()) {
TRI_ASSERT(false);
// this limit node affects only closest sort, if this sort is invalid
// we need to check other limit node
stopSearch = true;
}
}
#endif
if (stopSearch) {
// we have a doc body used before selected SortNode. Forget it, let`s look for better sort to use
sortNode = nullptr;
@ -314,7 +301,7 @@ void arangodb::aql::lateDocumentMaterializationRule(arangodb::aql::Optimizer* op
// on cluster we need to materialize node stay close to sort node on db server (to avoid network hop for materialization calls)
// however on single server we move it to limit node to make materialization as lazy as possible
auto materializeDependency = ServerState::instance()->isCoordinator() || stickToSortNode ? sortNode : limitNode;
auto materializeDependency = ServerState::instance()->isCoordinator() ? sortNode : limitNode;
auto dependencyParent = materializeDependency->getFirstParent();
TRI_ASSERT(dependencyParent != nullptr);
dependencyParent->replaceDependency(materializeDependency, materializeNode);

View File

@ -32,7 +32,7 @@ class Optimizer;
struct OptimizerRule;
class ExecutionPlan;
/// @brief moves document materialization from index nodes to materialize nodes
/// @brief moves document materialization from index nodes to sort nodes
void lateDocumentMaterializationRule(arangodb::aql::Optimizer* opt,
std::unique_ptr<arangodb::aql::ExecutionPlan> plan,
arangodb::aql::OptimizerRule const& rule);

View File

@ -211,24 +211,6 @@ function lateDocumentMaterializationArangoSearchRuleTestSuite () {
});
assertEqual(0, expected.size);
},
testQueryResultsWithBetweenCalc() {
let query = "FOR d IN " + svn + " SEARCH d.value IN [1,2, 11, 12] SORT BM25(d) LET c = NOOPT(CONCAT(d._key, '-C')) LIMIT 10 RETURN c ";
let plan = AQL_EXPLAIN(query).plan;
if (!isCluster) {
assertNotEqual(-1, plan.rules.indexOf(ruleName));
let result = AQL_EXECUTE(query);
assertEqual(4, result.json.length);
let expected = new Set(['c1-C', 'c2-C', 'c_1-C', 'c_2-C']);
result.json.forEach(function(doc) {
assertTrue(expected.has(doc));
expected.delete(doc);
});
assertEqual(0, expected.size);
} else {
// on cluster this will not be applied as calculation node will be moved up
assertEqual(-1, plan.rules.indexOf(ruleName));
}
},
testQueryResultsSkipSome() {
let query = "FOR d IN " + vn + " SEARCH PHRASE(d.str, 'cat', 'text_en') SORT TFIDF(d) DESC LIMIT 4, 1 RETURN d ";
let plan = AQL_EXPLAIN(query).plan;

View File

@ -279,26 +279,6 @@ function lateDocumentMaterializationRuleTestSuite () {
assertEqual(0, expected.size);
}
},
testQueryResultsWithBetweenCalc() {
for (i = 0; i < numOfCollectionIndexes; ++i) {
let query = "FOR d IN " + collectionNames[i] + " FILTER d.obj.a == 'a_val' SORT d.obj.c LET c = CONCAT(NOOPT(d.obj.d), '-C') LIMIT 10 RETURN c";
let plan = AQL_EXPLAIN(query).plan;
if (!isCluster) {
assertNotEqual(-1, plan.rules.indexOf(ruleName));
let result = AQL_EXECUTE(query);
assertEqual(2, result.json.length);
let expected = new Set(['d_val-C', 'd_val_2-C']);
result.json.forEach(function(doc) {
assertTrue(expected.has(doc));
expected.delete(doc);
});
assertEqual(0, expected.size);
} else {
// on cluster this will not be applied as calculation node will be moved up
assertEqual(-1, plan.rules.indexOf(ruleName));
}
}
},
testQueryResultsSkipSome() {
for (i = 0; i < numOfCollectionIndexes; ++i) {
let query = "FOR d IN " + collectionNames[i] + " FILTER d.obj.a == 'a_val' SORT d.obj.c DESC LIMIT 1, 1 RETURN d";