PostgreSQL 소스 판독(66) - 쿼리 문장 #51(make one rel 함수 #16-merge join#1)
1. 데이터 구조
Cost 관련 주의: 실제 사용하는 매개 변수 값은 시스템 설정 파일을 통해 정의됩니다. 이곳의 상수 정의가 아닙니다.
typedef double Cost; /* execution cost (in page-access units) */
/* defaults for costsize.c's Cost parameters */
/* NB: cost-estimation code should use the variables, not these constants! */
/* : , ! */
/* If you change these, update backend/utils/misc/postgresql.sample.conf */
#define DEFAULT_SEQ_PAGE_COST 1.0 // page
#define DEFAULT_RANDOM_PAGE_COST 4.0 // page
#define DEFAULT_CPU_TUPLE_COST 0.01 // CPU
#define DEFAULT_CPU_INDEX_TUPLE_COST 0.005 // CPU
#define DEFAULT_CPU_OPERATOR_COST 0.0025 // CPU
#define DEFAULT_PARALLEL_TUPLE_COST 0.1 // , worker worker
#define DEFAULT_PARALLEL_SETUP_COST 1000.0 //
#define DEFAULT_EFFECTIVE_CACHE_SIZE 524288 /* , measured in pages */
double seq_page_cost = DEFAULT_SEQ_PAGE_COST;
double random_page_cost = DEFAULT_RANDOM_PAGE_COST;
double cpu_tuple_cost = DEFAULT_CPU_TUPLE_COST;
double cpu_index_tuple_cost = DEFAULT_CPU_INDEX_TUPLE_COST;
double cpu_operator_cost = DEFAULT_CPU_OPERATOR_COST;
double parallel_tuple_cost = DEFAULT_PARALLEL_TUPLE_COST;
double parallel_setup_cost = DEFAULT_PARALLEL_SETUP_COST;
int effective_cache_size = DEFAULT_EFFECTIVE_CACHE_SIZE;
Cost disable_cost = 1.0e10;//1 10 0, ,
int max_parallel_workers_per_gather = 2;// gather worker
2. 원본 코드 해석
sort_inner_and_outer 함수는mergejoin 접근 경로를 구성하려고 시도합니다.구조 과정 중의 원가 추산 실현 함수 initialcost_mergejoin과finalcost_mergejoin은 다음 절에서 소개합니다.
//------------------------------------------------ sort_inner_and_outer
/*
* sort_inner_and_outer
* Create mergejoin join paths by explicitly sorting both the outer and
* inner join relations on each available merge ordering.
* outer inner , mergejoin .
*
* 'joinrel' is the join relation
* joinrel- relation
* 'outerrel' is the outer join relation
* outerrel- outer relation( , )
* 'innerrel' is the inner join relation
* innerrel- inner relation( )
* 'jointype' is the type of join to do
* jointype-
* 'extra' contains additional input values
* extra-
*/
static void
sort_inner_and_outer(PlannerInfo *root,
RelOptInfo *joinrel,
RelOptInfo *outerrel,
RelOptInfo *innerrel,
JoinType jointype,
JoinPathExtraData *extra)
{
JoinType save_jointype = jointype;
Path *outer_path;
Path *inner_path;
Path *cheapest_partial_outer = NULL;
Path *cheapest_safe_inner = NULL;
List *all_pathkeys;
ListCell *l;
/*
* We only consider the cheapest-total-cost input paths, since we are
* assuming here that a sort is required. We will consider
* cheapest-startup-cost input paths later, and only if they don't need a
* sort.
* , 。
* , 。
*
* This function intentionally does not consider parameterized input
* paths, except when the cheapest-total is parameterized. If we did so,
* we'd have a combinatorial explosion of mergejoin paths of dubious
* value. This interacts with decisions elsewhere that also discriminate
* against mergejoins with parameterized inputs; see comments in
* src/backend/optimizer/README.
* , 。
* ,mergejoin , 。
* , " " mergejoin;
* src/backend/optimizer/README 。
*/
outer_path = outerrel->cheapest_total_path;
inner_path = innerrel->cheapest_total_path;
/*
* If either cheapest-total path is parameterized by the other rel, we
* can't use a mergejoin. (There's no use looking for alternative input
* paths, since these should already be the least-parameterized available
* paths.)
* , 。
* ( , 。)
*/
if (PATH_PARAM_BY_REL(outer_path, innerrel) ||
PATH_PARAM_BY_REL(inner_path, outerrel))
return;
/*
* If unique-ification is requested, do it and then handle as a plain
* inner join.
* , JOIN_INNER
*/
if (jointype == JOIN_UNIQUE_OUTER)
{
outer_path = (Path *) create_unique_path(root, outerrel,
outer_path, extra->sjinfo);
Assert(outer_path);
jointype = JOIN_INNER;
}
else if (jointype == JOIN_UNIQUE_INNER)
{
inner_path = (Path *) create_unique_path(root, innerrel,
inner_path, extra->sjinfo);
Assert(inner_path);
jointype = JOIN_INNER;
}
/*
* If the joinrel is parallel-safe, we may be able to consider a partial
* merge join. However, we can't handle JOIN_UNIQUE_OUTER, because the
* outer path will be partial, and therefore we won't be able to properly
* guarantee uniqueness. Similarly, we can't handle JOIN_FULL and
* JOIN_RIGHT, because they can produce false null extended rows. Also,
* the resulting path must not be parameterized.
* , 。
* ,PG JOIN_UNIQUE_OUTER, , 。
* ,PG JOIN_FULL JOIN_RIGHT, 。
* , 。
*/
if (joinrel->consider_parallel &&
save_jointype != JOIN_UNIQUE_OUTER &&
save_jointype != JOIN_FULL &&
save_jointype != JOIN_RIGHT &&
outerrel->partial_pathlist != NIL &&
bms_is_empty(joinrel->lateral_relids))
{
cheapest_partial_outer = (Path *) linitial(outerrel->partial_pathlist);
if (inner_path->parallel_safe)
cheapest_safe_inner = inner_path;
else if (save_jointype != JOIN_UNIQUE_INNER)
cheapest_safe_inner =
get_cheapest_parallel_safe_total_inner(innerrel->pathlist);
}
/*
* Each possible ordering of the available mergejoin clauses will generate
* a differently-sorted result path at essentially the same cost. We have
* no basis for choosing one over another at this level of joining, but
* some sort orders may be more useful than others for higher-level
* mergejoins, so it's worth considering multiple orderings.
* mergejoin 。
* , , ,
* , 。
*
* Actually, it's not quite true that every mergeclause ordering will
* generate a different path order, because some of the clauses may be
* partially redundant (refer to the same EquivalenceClasses). Therefore,
* what we do is convert the mergeclause list to a list of canonical
* pathkeys, and then consider different orderings of the pathkeys.
* , mergeclause , ( )。
* , mergeclause , 。
*
* Generating a path for *every* permutation of the pathkeys doesn't seem
* like a winning strategy; the cost in planning time is too high. For
* now, we generate one path for each pathkey, listing that pathkey first
* and the rest in random order. This should allow at least a one-clause
* mergejoin without re-sorting against any other possible mergejoin
* partner path. But if we've not guessed the right ordering of secondary
* keys, we may end up evaluating clauses as qpquals when they could have
* been done as mergeclauses. (In practice, it's rare that there's more
* than two or three mergeclauses, so expending a huge amount of thought
* on that is probably not worth it.)
* , 。
* , pathkey , pathkey, 。
* mergejoin, mergejoin 。
* , qpquals , mergeclauses 。
* ( , mergeclauses, )。
*
* The pathkey order returned by select_outer_pathkeys_for_merge() has
* some heuristics behind it (see that function), so be sure to try it
* exactly as-is as well as making variants.
* select_outer_pathkeys_for_merge pathkey
* ( ), , 。
*/
all_pathkeys = select_outer_pathkeys_for_merge(root,
extra->mergeclause_list,
joinrel);
foreach(l, all_pathkeys)//
{
List *front_pathkey = (List *) lfirst(l);
List *cur_mergeclauses;
List *outerkeys;
List *innerkeys;
List *merge_pathkeys;
/* Make a pathkey list with this guy first */
if (l != list_head(all_pathkeys))
outerkeys = lcons(front_pathkey,
list_delete_ptr(list_copy(all_pathkeys),
front_pathkey));
else
outerkeys = all_pathkeys; /* no work at first one... */
/* Sort the mergeclauses into the corresponding ordering */
cur_mergeclauses =
find_mergeclauses_for_outer_pathkeys(root,
outerkeys,
extra->mergeclause_list);
/* Should have used them all... */
Assert(list_length(cur_mergeclauses) == list_length(extra->mergeclause_list));
/* Build sort pathkeys for the inner side */
innerkeys = make_inner_pathkeys_for_merge(root,
cur_mergeclauses,
outerkeys);
/* Build pathkeys representing output sort order */
merge_pathkeys = build_join_pathkeys(root, joinrel, jointype,
outerkeys);
/*
* And now we can make the path.
*
* Note: it's possible that the cheapest paths will already be sorted
* properly. try_mergejoin_path will detect that case and suppress an
* explicit sort step, so we needn't do so here.
* : 。
* try_mergejoin_path , 。
*/
try_mergejoin_path(root,
joinrel,
outer_path,
inner_path,
merge_pathkeys,
cur_mergeclauses,
outerkeys,
innerkeys,
jointype,
extra,
false);
/*
* If we have partial outer and parallel safe inner path then try
* partial mergejoin path.
*
*/
if (cheapest_partial_outer && cheapest_safe_inner)
try_partial_mergejoin_path(root,
joinrel,
cheapest_partial_outer,
cheapest_safe_inner,
merge_pathkeys,
cur_mergeclauses,
outerkeys,
innerkeys,
jointype,
extra);
}
}
//----------------------------------- try_mergejoin_path
/*
* try_mergejoin_path
* Consider a merge join path; if it appears useful, push it into
* the joinrel's pathlist via add_path().
* merge join path, , add_path joinrel pathlist
*/
static void
try_mergejoin_path(PlannerInfo *root,
RelOptInfo *joinrel,
Path *outer_path,
Path *inner_path,
List *pathkeys,
List *mergeclauses,
List *outersortkeys,
List *innersortkeys,
JoinType jointype,
JoinPathExtraData *extra,
bool is_partial)
{
Relids required_outer;
JoinCostWorkspace workspace;
if (is_partial)
{
try_partial_mergejoin_path(root,
joinrel,
outer_path,
inner_path,
pathkeys,
mergeclauses,
outersortkeys,
innersortkeys,
jointype,
extra);//
return;
}
/*
* Check to see if proposed path is still parameterized, and reject if the
* parameterization wouldn't be sensible.
* , , 。
*/
required_outer = calc_non_nestloop_required_outer(outer_path,
inner_path);
if (required_outer &&
!bms_overlap(required_outer, extra->param_source_rels))
{
/* Waste no memory when we reject a path here */
bms_free(required_outer);
return;
}
/*
* If the given paths are already well enough ordered, we can skip doing
* an explicit sort.
* , .
*/
if (outersortkeys &&
pathkeys_contained_in(outersortkeys, outer_path->pathkeys))
outersortkeys = NIL;
if (innersortkeys &&
pathkeys_contained_in(innersortkeys, inner_path->pathkeys))
innersortkeys = NIL;
/*
* See comments in try_nestloop_path().
*/
initial_cost_mergejoin(root, &workspace, jointype, mergeclauses,
outer_path, inner_path,
outersortkeys, innersortkeys,
extra);// mergejoin
if (add_path_precheck(joinrel,
workspace.startup_cost, workspace.total_cost,
pathkeys, required_outer))//
{
add_path(joinrel, (Path *)
create_mergejoin_path(root,
joinrel,
jointype,
&workspace,
extra,
outer_path,
inner_path,
extra->restrictlist,
pathkeys,
required_outer,
mergeclauses,
outersortkeys,
innersortkeys));//
}
else
{
/* Waste no memory when we reject a path here */
bms_free(required_outer);
}
}
//----------------------- create_mergejoin_path
/*
* create_mergejoin_path
* Creates a pathnode corresponding to a mergejoin join between
* two relations
* merge join
*
* 'joinrel' is the join relation
* joinrel- relation
* 'jointype' is the type of join required
* jointype-
* 'workspace' is the result from initial_cost_mergejoin
* workspace- initial_cost_mergejoin
* 'extra' contains various information about the join
* extra-
* 'outer_path' is the outer path
* outer_path-outer relation
* 'inner_path' is the inner path
* inner_path-inner relation
* 'restrict_clauses' are the RestrictInfo nodes to apply at the join
* restrict_clauses-
* 'pathkeys' are the path keys of the new join path
* pathkeys-
* 'required_outer' is the set of required outer rels
* required_outer- outer rels, relids
* 'mergeclauses' are the RestrictInfo nodes to use as merge clauses
* (this should be a subset of the restrict_clauses list)
* mergeclauses-
* 'outersortkeys' are the sort varkeys for the outer relation
* outersortkeys-outer relation
* 'innersortkeys' are the sort varkeys for the inner relation
* innersortkeys-inner relation
*/
MergePath *
create_mergejoin_path(PlannerInfo *root,
RelOptInfo *joinrel,
JoinType jointype,
JoinCostWorkspace *workspace,
JoinPathExtraData *extra,
Path *outer_path,
Path *inner_path,
List *restrict_clauses,
List *pathkeys,
Relids required_outer,
List *mergeclauses,
List *outersortkeys,
List *innersortkeys)
{
MergePath *pathnode = makeNode(MergePath);
pathnode->jpath.path.pathtype = T_MergeJoin;
pathnode->jpath.path.parent = joinrel;
pathnode->jpath.path.pathtarget = joinrel->reltarget;
pathnode->jpath.path.param_info =
get_joinrel_parampathinfo(root,
joinrel,
outer_path,
inner_path,
extra->sjinfo,
required_outer,
&restrict_clauses);
pathnode->jpath.path.parallel_aware = false;
pathnode->jpath.path.parallel_safe = joinrel->consider_parallel &&
outer_path->parallel_safe && inner_path->parallel_safe;
/* This is a foolish way to estimate parallel_workers, but for now... */
pathnode->jpath.path.parallel_workers = outer_path->parallel_workers;
pathnode->jpath.path.pathkeys = pathkeys;
pathnode->jpath.jointype = jointype;
pathnode->jpath.inner_unique = extra->inner_unique;
pathnode->jpath.outerjoinpath = outer_path;
pathnode->jpath.innerjoinpath = inner_path;
pathnode->jpath.joinrestrictinfo = restrict_clauses;
pathnode->path_mergeclauses = mergeclauses;
pathnode->outersortkeys = outersortkeys;
pathnode->innersortkeys = innersortkeys;
/* pathnode->skip_mark_restore will be set by final_cost_mergejoin */
/* pathnode->materialize_inner will be set by final_cost_mergejoin */
final_cost_mergejoin(root, pathnode, workspace, extra);//
return pathnode;
}
3. 추적 분석
테스트 스크립트는 다음과 같습니다
select a.*,b.grbh,b.je
from t_dwxx a,
lateral (select t1.dwbh,t1.grbh,t2.je
from t_grxx t1
inner join t_jfxx t2 on t1.dwbh = a.dwbh and t1.grbh = t2.grbh) b
order by b.dwbh;
gdb 시작, 인터럽트 설정
(gdb) b sort_inner_and_outer
Breakpoint 1 at 0x7af63a: file joinpath.c, line 888.
(gdb) c
Continuing.
Breakpoint 1, sort_inner_and_outer (root=0x1a4a278, joinrel=0x1aa7180, outerrel=0x1a55700, innerrel=0x1a56c30,
jointype=JOIN_INNER, extra=0x7ffca933f880) at joinpath.c:888
888 JoinType save_jointype = jointype;
(gdb)
새로 생성된joinrel은 1번과 3번 RTE의 연결이며, 유형은 JOININNER
(gdb) p *joinrel->relids->words
$1 = 10
(gdb) p jointype
$2 = JOIN_INNER
정렬 키 가져오기, PathKey의 등가 클래스 EC, 구성원 tgrxx.dwbh 및 tdwxx.dwbh
...
(gdb)
993 all_pathkeys = select_outer_pathkeys_for_merge(root,
(gdb) n
997 foreach(l, all_pathkeys)
(gdb) p *all_pathkeys
$3 = {type = T_List, length = 1, head = 0x1a69490, tail = 0x1a69490}
(gdb) p *(PathKey *)all_pathkeys->head->data.ptr_value
$5 = {type = T_PathKey, pk_eclass = 0x1a60e08, pk_opfamily = 1994, pk_strategy = 1, pk_nulls_first = false}
...
(gdb) set $rt=(RelabelType *)((EquivalenceMember *)$ec->ec_members->head->data.ptr_value)->em_expr
(gdb) p *$rt->arg
$14 = {type = T_Var}
(gdb) p *(Var *)$rt->arg
$15 = {xpr = {type = T_Var}, varno = 3, varattno = 1, vartype = 1043, vartypmod = 14, varcollid = 100, varlevelsup = 0,
varnoold = 3, varoattno = 1, location = 208}
(gdb) set $rt2=(RelabelType *)((EquivalenceMember *)$ec->ec_members->head->next->data.ptr_value)->em_expr
(gdb) p *(Var *)$rt2->arg
$16 = {xpr = {type = T_Var}, varno = 1, varattno = 2, vartype = 1043, vartypmod = 24, varcollid = 100, varlevelsup = 0,
varnoold = 1, varoattno = 2, location = 218}
시작allpathkeys
(gdb) n
999 List *front_pathkey = (List *) lfirst(l);
연결 조건 서브문장 가져오기, tdwxx.dwbh=t_grxx.dwbh
(gdb) p *cur_mergeclauses
$17 = {type = T_List, length = 1, head = 0x1a694f0, tail = 0x1a694f0}
outer와 inner relation의 정렬 키 만들기
(gdb) p *(PathKey *)innerkeys->head->data.ptr_value
$22 = {type = T_PathKey, pk_eclass = 0x1a60e08, pk_opfamily = 1994, pk_strategy = 1, pk_nulls_first = false}
(gdb) p *(PathKey *)merge_pathkeys->head->data.ptr_value
$25 = {type = T_PathKey, pk_eclass = 0x1a60e08, pk_opfamily = 1994, pk_strategy = 1, pk_nulls_first = false}
merge join 시도, 함수try 진입mergejoin_path
(gdb)
1038 try_mergejoin_path(root,
(gdb) step
try_mergejoin_path (root=0x1a4dcc0, joinrel=0x1a68e20, outer_path=0x1a62288, inner_path=0x1a62320, pathkeys=0x1a694b8,
mergeclauses=0x1a69518, outersortkeys=0x1a694b8, innersortkeys=0x1a69578, jointype=JOIN_INNER, extra=0x7ffca933f880,
is_partial=false) at joinpath.c:572
572 if (is_partial)
초기 mergejoin 비용
...
(gdb)
615 initial_cost_mergejoin(root, &workspace, jointype, mergeclauses,
(gdb) p workspace
$26 = {startup_cost = 10861.483356195882, total_cost = 11134.203356195881, run_cost = 24.997499999999999,
inner_run_cost = 247.72250000000003, inner_rescan_run_cost = 1.3627136827435593e-316, outer_rows = 9999,
inner_rows = 100000, outer_skip_rows = 0, inner_skip_rows = 911, numbuckets = 27665584, numbatches = 0,
inner_rows_total = 1.3681950446447804e-316}
구조mergejoin
...
(gdb) n
625 create_mergejoin_path(root,
(gdb)
624 add_path(joinrel, (Path *)
(gdb)
644 }
(gdb) p *joinrel->pathlist
$28 = {type = T_List, length = 1, head = 0x1a6a180, tail = 0x1a6a180}
(gdb) p *(Node *)joinrel->pathlist->head->data.ptr_value
$29 = {type = T_MergePath}
(gdb) p *(MergePath *)joinrel->pathlist->head->data.ptr_value
$30 = {jpath = {path = {type = T_MergePath, pathtype = T_MergeJoin, parent = 0x1a68e20, pathtarget = 0x1a69058,
param_info = 0x0, parallel_aware = false, parallel_safe = true, parallel_workers = 0, rows = 100000,
startup_cost = 10863.760856195882, total_cost = 12409.200856195883, pathkeys = 0x1a694b8}, jointype = JOIN_INNER,
inner_unique = false, outerjoinpath = 0x1a62288, innerjoinpath = 0x1a62320, joinrestrictinfo = 0x1a692f8},
path_mergeclauses = 0x1a69518, outersortkeys = 0x1a694b8, innersortkeys = 0x1a69578, skip_mark_restore = false,
materialize_inner = false}
호출 완료
(gdb) n
sort_inner_and_outer (root=0x1a4dcc0, joinrel=0x1a68e20, outerrel=0x1a4d700, innerrel=0x1a4d918, jointype=JOIN_INNER,
extra=0x7ffca933f880) at joinpath.c:1054
1054 if (cheapest_partial_outer && cheapest_safe_inner)
(gdb)
997 foreach(l, all_pathkeys)
(gdb)
1066 }
(gdb) n
add_paths_to_joinrel (root=0x1a4dcc0, joinrel=0x1a68e20, outerrel=0x1a4d700, innerrel=0x1a4d918, jointype=JOIN_INNER,
sjinfo=0x7ffca933f970, restrictlist=0x1a692f8) at joinpath.c:279
279 if (mergejoin_allowed)
(gdb)
280 match_unsorted_outer(root, joinrel, outerrel, innerrel,
...
DONE!
4. 참고 자료
allpaths.c cost.h costsize.c PG Document:Query Planning
이 내용에 흥미가 있습니까?
현재 기사가 여러분의 문제를 해결하지 못하는 경우 AI 엔진은 머신러닝 분석(스마트 모델이 방금 만들어져 부정확한 경우가 있을 수 있음)을 통해 가장 유사한 기사를 추천합니다:
다양한 언어의 JSONJSON은 Javascript 표기법을 사용하여 데이터 구조를 레이아웃하는 데이터 형식입니다. 그러나 Javascript가 코드에서 이러한 구조를 나타낼 수 있는 유일한 언어는 아닙니다. 저는 일반적으로 '객체'{}...
텍스트를 자유롭게 공유하거나 복사할 수 있습니다.하지만 이 문서의 URL은 참조 URL로 남겨 두십시오.
CC BY-SA 2.5, CC BY-SA 3.0 및 CC BY-SA 4.0에 따라 라이센스가 부여됩니다.