hcq/oceanbase
hcq/oceanbase
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[CP] Fix limit group by cardinality estimation bug

Browse Source
This commit is contained in:
xianyu-w 2024-07-11 13:17:10 +00:00 committed by ob-robot
parent f73d4736a1
commit 0e9c4cda2b
3 changed files with 46 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

42
src/sql/optimizer/ob_log_distinct.cpp
View File

@ -169,7 +169,15 @@ int ObLogDistinct::do_re_est_cost(EstimateCostInfo &param, double &card, double
param.need_row_count_ < child_card &&
param.need_row_count_ < total_ndv_) {
child_ndv = param.need_row_count_;
param.need_row_count_ = child_card * (1 - std::pow((1 - child_ndv / total_ndv_), total_ndv_ / child_card));
if (input_sorted_) {
if (param.need_row_count_ > 1.0) {
param.need_row_count_ = child_card * (param.need_row_count_ - 1.0) / total_ndv_ + 1.0;
} else {
// do nothing
}
} else {
param.need_row_count_ = child_card * (1 - std::pow((1 - child_ndv / total_ndv_), total_ndv_ / child_card));
}
} else {
param.need_row_count_ = -1;
need_scale_ndv = true;
@ -395,5 +403,37 @@ int ObLogDistinct::check_use_child_ordering(bool &used, int64_t &inherit_child_o
return ret;
}
int ObLogDistinct::compute_property()
{
int ret = OB_SUCCESS;
ObLogicalOperator *top = get_child(first_child);
bool need_sort = false;
int64_t prefix_pos = 0;
if (OB_ISNULL(top) || OB_ISNULL(get_plan())) {
ret = OB_ERR_UNEXPECTED;
LOG_WARN("unexpected null", K(ret));
} else if (OB_FAIL(ObLogicalOperator::compute_property())) {
LOG_WARN("failed to compute op property", K(ret));
} else if (MERGE_AGGREGATE == algo_) {
input_sorted_ = true;
} else if (OB_FAIL(ObOptimizerUtil::check_need_sort(distinct_exprs_,
NULL/*directions*/,
top->get_op_ordering(),
top->get_fd_item_set(),
top->get_output_equal_sets(),
top->get_output_const_exprs(),
get_plan()->get_onetime_query_refs(),
top->get_is_at_most_one_row(),
need_sort,
prefix_pos))) {
LOG_WARN("failed to check need sort", K(ret));
} else if (!need_sort) {
input_sorted_ = true;
} else {
input_sorted_ = false;
}
return ret;
}
}
}

5
src/sql/optimizer/ob_log_distinct.h
View File

@ -28,7 +28,8 @@ public:
is_block_mode_(false),
is_push_down_(false),
total_ndv_(-1.0),
force_push_down_(false)
force_push_down_(false),
input_sorted_(false)
{ }
virtual ~ObLogDistinct()
{ }
@ -76,6 +77,7 @@ public:
inline void set_is_partition_gi(bool v) { is_partition_gi_ = v; }
virtual int get_card_without_filter(double &card) override;
virtual int check_use_child_ordering(bool &used, int64_t &inherit_child_ordering_index)override;
virtual int compute_property() override;
private:
common::ObSEArray<ObRawExpr*, 16, common::ModulePageAllocator, true> distinct_exprs_;
@ -85,6 +87,7 @@ private:
double total_ndv_;
bool force_push_down_; // control by _aggregation_optimization_settings
bool is_partition_gi_;
bool input_sorted_;
private:
DISALLOW_COPY_AND_ASSIGN(ObLogDistinct);
};

2
src/sql/optimizer/ob_log_group_by.cpp
View File

@ -285,7 +285,7 @@ int ObLogGroupBy::do_re_est_cost(EstimateCostInfo &param, double &card, double &
need_ndv /= selectivity;
}
if (child_card > 0) {
param.need_row_count_ = child_card * (1 - std::pow((1 - need_ndv / child_ndv), child_ndv / child_card));
param.need_row_count_ = child_card * need_ndv / child_ndv;
param.need_row_count_ /= number_of_copies;
} else {
param.need_row_count_ = 0;