You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@quickstep.apache.org by zu...@apache.org on 2017/01/26 00:43:08 UTC
[2/2] incubator-quickstep git commit: Added limited optimizer support
for Partitioned Hash Joins.
Added limited optimizer support for Partitioned Hash Joins.
Project: http://git-wip-us.apache.org/repos/asf/incubator-quickstep/repo
Commit: http://git-wip-us.apache.org/repos/asf/incubator-quickstep/commit/71328511
Tree: http://git-wip-us.apache.org/repos/asf/incubator-quickstep/tree/71328511
Diff: http://git-wip-us.apache.org/repos/asf/incubator-quickstep/diff/71328511
Branch: refs/heads/reorder-partitioned-hash-join
Commit: 71328511403eeb9f744193244548a7f75fa48ffa
Parents: 968ce3f
Author: Zuyu Zhang <zu...@apache.org>
Authored: Wed Jan 25 01:49:28 2017 -0800
Committer: Zuyu Zhang <zu...@apache.org>
Committed: Wed Jan 25 16:42:50 2017 -0800
----------------------------------------------------------------------
query_optimizer/ExecutionGenerator.cpp | 149 +++++++++++++++----
.../tests/execution_generator/Partition.test | 51 +++++++
types/TypedValue.hpp | 19 +++
3 files changed, 190 insertions(+), 29 deletions(-)
----------------------------------------------------------------------
http://git-wip-us.apache.org/repos/asf/incubator-quickstep/blob/71328511/query_optimizer/ExecutionGenerator.cpp
----------------------------------------------------------------------
diff --git a/query_optimizer/ExecutionGenerator.cpp b/query_optimizer/ExecutionGenerator.cpp
index e25b8ad..5035b4b 100644
--- a/query_optimizer/ExecutionGenerator.cpp
+++ b/query_optimizer/ExecutionGenerator.cpp
@@ -132,8 +132,10 @@
#include "gflags/gflags.h"
#include "glog/logging.h"
+using std::find;
using std::move;
using std::static_pointer_cast;
+using std::swap;
using std::unique_ptr;
using std::unordered_map;
using std::vector;
@@ -608,6 +610,44 @@ void ExecutionGenerator::convertSharedSubplanReference(const physical::SharedSub
}
}
+namespace {
+
+bool areSamePartitionSchemeHeaders(const PartitionSchemeHeader &lhs_partition_header,
+ const CatalogRelationSchema &lhs_scheme,
+ const PartitionSchemeHeader &rhs_partition_header,
+ const CatalogRelationSchema &rhs_scheme) {
+ if (lhs_partition_header.getPartitionType() != rhs_partition_header.getPartitionType()) {
+ return false;
+ }
+
+ if (lhs_partition_header.getNumPartitions() != rhs_partition_header.getNumPartitions()) {
+ return false;
+ }
+
+ // Check whether the underlying types in CatalogAttribute are the same.
+ if (!lhs_scheme.getAttributeById(lhs_partition_header.getPartitionAttributeId())->getType().equals(
+ rhs_scheme.getAttributeById(rhs_partition_header.getPartitionAttributeId())->getType())) {
+ return false;
+ }
+
+ switch (lhs_partition_header.getPartitionType()) {
+ case PartitionSchemeHeader::PartitionType::kHash:
+ return true;
+ case PartitionSchemeHeader::PartitionType::kRange: {
+ const vector<TypedValue> &lhs_ranges =
+ static_cast<const RangePartitionSchemeHeader&>(lhs_partition_header).getPartitionRangeBoundaries();
+ const vector<TypedValue> &rhs_ranges =
+ static_cast<const RangePartitionSchemeHeader&>(rhs_partition_header).getPartitionRangeBoundaries();
+
+ return lhs_ranges == rhs_ranges;
+ }
+ }
+
+ return false;
+}
+
+} // namespace
+
void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
// HashJoin is converted to three operators:
// BuildHash, HashJoin, DestroyHash. The second is the primary operator.
@@ -618,13 +658,10 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
std::vector<attribute_id> probe_attribute_ids;
std::vector<attribute_id> build_attribute_ids;
- std::size_t build_cardinality =
- cost_model_for_hash_join_->estimateCardinality(build_physical);
-
bool any_probe_attributes_nullable = false;
bool any_build_attributes_nullable = false;
- const std::vector<E::AttributeReferencePtr> &left_join_attributes =
+ std::vector<E::AttributeReferencePtr> left_join_attributes =
physical_plan->left_join_attributes();
for (const E::AttributeReferencePtr &left_join_attribute : left_join_attributes) {
const CatalogAttribute *probe_catalog_attribute
@@ -636,7 +673,7 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
}
}
- const std::vector<E::AttributeReferencePtr> &right_join_attributes =
+ std::vector<E::AttributeReferencePtr> right_join_attributes =
physical_plan->right_join_attributes();
for (const E::AttributeReferencePtr &right_join_attribute : right_join_attributes) {
const CatalogAttribute *build_catalog_attribute
@@ -648,6 +685,76 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
}
}
+ const CatalogRelationInfo *build_relation_info =
+ findRelationInfoOutputByPhysical(build_physical);
+ const CatalogRelationInfo *probe_operator_info =
+ findRelationInfoOutputByPhysical(probe_physical);
+
+ const CatalogRelation *build_relation = build_relation_info->relation;
+ const CatalogRelation *probe_relation = probe_operator_info->relation;
+
+ const PartitionScheme *build_partition_scheme = build_relation->getPartitionScheme();
+ const PartitionScheme *probe_partition_scheme = probe_relation->getPartitionScheme();
+
+ // Check whether we need repartition for both build and probe.
+ if (build_partition_scheme) {
+ const PartitionSchemeHeader &build_partition_scheme_header = build_partition_scheme->getPartitionSchemeHeader();
+ switch (build_partition_scheme_header.getPartitionType()) {
+ case PartitionSchemeHeader::PartitionType::kRange:
+ THROW_SQL_ERROR() << "Re-partition for range-based is not supported";
+ case PartitionSchemeHeader::PartitionType::kHash: {
+ const attribute_id build_partition_attr = build_partition_scheme_header.getPartitionAttributeId();
+ if (find(build_attribute_ids.begin(), build_attribute_ids.end(), build_partition_attr) ==
+ build_attribute_ids.end()) {
+ THROW_SQL_ERROR() << "Re-partition for hash-based is not supported";
+ }
+ break;
+ }
+ }
+ }
+
+ if (probe_partition_scheme) {
+ const PartitionSchemeHeader &probe_partition_scheme_header = probe_partition_scheme->getPartitionSchemeHeader();
+ switch (probe_partition_scheme_header.getPartitionType()) {
+ case PartitionSchemeHeader::PartitionType::kRange:
+ THROW_SQL_ERROR() << "Re-partition for range-based is not supported";
+ case PartitionSchemeHeader::PartitionType::kHash: {
+ const attribute_id probe_partition_attr = probe_partition_scheme_header.getPartitionAttributeId();
+ if (find(probe_attribute_ids.begin(), probe_attribute_ids.end(), probe_partition_attr) ==
+ probe_attribute_ids.end()) {
+ THROW_SQL_ERROR() << "Re-partition for hash-based is not supported";
+ }
+ break;
+ }
+ }
+ }
+
+ // No partition.
+ std::size_t build_num_partitions = 1;
+
+ // Set 'build_num_partitions', and swap the build and probe, if needed.
+ if (build_partition_scheme && probe_partition_scheme) {
+ const PartitionSchemeHeader &build_partition_scheme_header = build_partition_scheme->getPartitionSchemeHeader();
+ const PartitionSchemeHeader &probe_partition_scheme_header = probe_partition_scheme->getPartitionSchemeHeader();
+ if (!areSamePartitionSchemeHeaders(build_partition_scheme_header, *build_relation,
+ probe_partition_scheme_header, *probe_relation)) {
+ THROW_SQL_ERROR() << "Hash Join of two different HashPartitionSchemeHeaders is not supported";
+ }
+
+ build_num_partitions = build_partition_scheme->getPartitionSchemeHeader().getNumPartitions();
+ } else if (build_partition_scheme) {
+ build_num_partitions = build_partition_scheme->getPartitionSchemeHeader().getNumPartitions();
+ } else if (probe_partition_scheme) {
+ build_num_partitions = probe_partition_scheme->getPartitionSchemeHeader().getNumPartitions();
+
+ swap(probe_physical, build_physical);
+ swap(probe_attribute_ids, build_attribute_ids);
+ swap(any_probe_attributes_nullable, any_build_attributes_nullable);
+ swap(left_join_attributes, right_join_attributes);
+ swap(probe_operator_info, build_relation_info);
+ swap(probe_relation, build_relation);
+ }
+
// Remember key types for call to SimplifyHashTableImplTypeProto() below.
std::vector<const Type*> key_types;
for (std::vector<E::AttributeReferencePtr>::size_type attr_idx = 0;
@@ -677,11 +784,6 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
convertNamedExpressions(physical_plan->project_expressions(),
query_context_proto_->add_scalar_groups());
- const CatalogRelationInfo *build_relation_info =
- findRelationInfoOutputByPhysical(build_physical);
- const CatalogRelationInfo *probe_operator_info =
- findRelationInfoOutputByPhysical(probe_physical);
-
// Create a vector that indicates whether each project expression is using
// attributes from the build relation as input. This information is required
// by the current implementation of hash left outer join
@@ -694,10 +796,8 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
build_physical->getOutputAttributes())));
}
- const CatalogRelation *build_relation = build_relation_info->relation;
-
// FIXME(quickstep-team): Add support for self-join.
- if (build_relation == probe_operator_info->relation) {
+ if (build_relation == probe_relation) {
THROW_SQL_ERROR() << "Self-join is not supported";
}
@@ -706,18 +806,7 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
query_context_proto_->join_hash_tables_size();
S::QueryContext::HashTableContext *hash_table_context_proto =
query_context_proto_->add_join_hash_tables();
-
- // No partition.
- std::size_t num_partitions = 1;
- if (build_relation->hasPartitionScheme() &&
- build_attribute_ids.size() == 1) {
- const PartitionSchemeHeader &partition_scheme_header =
- build_relation->getPartitionScheme()->getPartitionSchemeHeader();
- if (build_attribute_ids[0] == partition_scheme_header.getPartitionAttributeId()) {
- // TODO(zuyu): add optimizer support for partitioned hash joins.
- hash_table_context_proto->set_num_partitions(num_partitions);
- }
- }
+ hash_table_context_proto->set_num_partitions(build_num_partitions);
S::HashTable *hash_table_proto = hash_table_context_proto->mutable_join_hash_table();
@@ -734,6 +823,8 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
build_relation->getAttributeById(build_attribute)->getType().getProto());
}
+ const std::size_t build_cardinality =
+ cost_model_for_hash_join_->estimateCardinality(build_physical);
hash_table_proto->set_estimated_num_entries(build_cardinality);
// Create three operators.
@@ -745,7 +836,7 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
build_relation_info->isStoredRelation(),
build_attribute_ids,
any_build_attributes_nullable,
- num_partitions,
+ build_num_partitions,
join_hash_table_index));
// Create InsertDestination proto.
@@ -784,11 +875,11 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
new HashJoinOperator(
query_handle_->query_id(),
*build_relation,
- *probe_operator_info->relation,
+ *probe_relation,
probe_operator_info->isStoredRelation(),
probe_attribute_ids,
any_probe_attributes_nullable,
- num_partitions,
+ build_num_partitions,
*output_relation,
insert_destination_index,
join_hash_table_index,
@@ -800,7 +891,7 @@ void ExecutionGenerator::convertHashJoin(const P::HashJoinPtr &physical_plan) {
const QueryPlan::DAGNodeIndex destroy_operator_index =
execution_plan_->addRelationalOperator(new DestroyHashOperator(
- query_handle_->query_id(), num_partitions, join_hash_table_index));
+ query_handle_->query_id(), build_num_partitions, join_hash_table_index));
if (!build_relation_info->isStoredRelation()) {
execution_plan_->addDirectDependency(build_operator_index,
http://git-wip-us.apache.org/repos/asf/incubator-quickstep/blob/71328511/query_optimizer/tests/execution_generator/Partition.test
----------------------------------------------------------------------
diff --git a/query_optimizer/tests/execution_generator/Partition.test b/query_optimizer/tests/execution_generator/Partition.test
index ab05391..8e94316 100644
--- a/query_optimizer/tests/execution_generator/Partition.test
+++ b/query_optimizer/tests/execution_generator/Partition.test
@@ -18,12 +18,20 @@
CREATE TABLE foo (id INT NULL,
name CHAR(20))
PARTITION BY HASH(id) PARTITIONS 4;
+CREATE TABLE bar (id INT NULL,
+ score DOUBLE NULL)
+PARTITION BY HASH(id) PARTITIONS 4;
INSERT INTO foo
SELECT int_col, char_col
FROM test
WHERE int_col > 0 OR int_col < 0;
+INSERT INTO bar
+SELECT int_col, double_col
+FROM test
+WHERE int_col % 2 = 0;
+
SELECT * FROM foo;
--
+-----------+--------------------+
@@ -52,3 +60,46 @@ SELECT * FROM foo;
| -17| -17 4.123106|
| -21| -21 4.582576|
+-----------+--------------------+
+==
+
+# Partitioned Hash Join.
+SELECT bar.id, foo.name AS "char_col"
+FROM foo JOIN bar ON foo.id = bar.id;
+--
++-----------+--------------------+
+|id |char_col |
++-----------+--------------------+
+| 4| 4 2.000000|
+| 8| 8 2.828427|
+| 12| 12 3.464102|
+| 16| 16 4.000000|
+| 24| 24 4.898979|
+| 2| 2 1.414214|
+| 6| 6 2.449490|
+| 14| 14 3.741657|
+| 18| 18 4.242641|
+| 22| 22 4.690416|
++-----------+--------------------+
+==
+
+# Hash Join with two stored relations, one of which is partitioned.
+SELECT bar.id, test.char_col
+FROM test JOIN bar ON test.int_col = bar.id;
+--
+[same as above]
+==
+
+# Same as above, but has a different join order.
+SELECT bar.id, test.char_col
+FROM bar JOIN test ON bar.id = test.int_col;
+--
+[same as above]
+==
+
+# Hash Join with one stored, partitioned relation,
+# and a non-stored, non-partitioned one.
+SELECT bar.id, test.char_col
+FROM bar JOIN test ON bar.id = test.int_col
+WHERE test.int_col % 2 = 0;
+--
+[same as above]
http://git-wip-us.apache.org/repos/asf/incubator-quickstep/blob/71328511/types/TypedValue.hpp
----------------------------------------------------------------------
diff --git a/types/TypedValue.hpp b/types/TypedValue.hpp
index d75720a..832690d 100644
--- a/types/TypedValue.hpp
+++ b/types/TypedValue.hpp
@@ -253,6 +253,25 @@ class TypedValue {
}
/**
+ * @brief Equal operator.
+ **/
+ bool operator==(const TypedValue &rhs) const {
+ if (getTypeID() != rhs.getTypeID()) {
+ return false;
+ }
+
+ if (isNull() != rhs.isNull()) {
+ return false;
+ }
+
+ if (isNull()) {
+ return true;
+ }
+
+ return fastEqualCheck(rhs);
+ }
+
+ /**
* @brief Create a new literal TypedValue with pre-allocated out-of-line
* data.
* @warning The memory at value_ptr must be allocated with malloc() or