<html lang="en_US">
<head>
<title>pg_hint_plan</title>
-<!-- Uncoment after the tool has been hosted somewhere.
+<!-- Uncomment after the tool has been hosted somewhere.
<link rel="home" title="pg_hint_plan" href="index.html">
-->
<link rel="stylesheet" type="text/css" href="style.css">
</head>
<body>
-<h1 id="pg_hint_plan">pg_hint_plan 1.3</h1>
+<h1 id="pg_hint_plan">pg_hint_plan 1.6</h1>
<div class="navigation">
- <a href="pg_hint_plan-ja.html">pg_hint_plan</a>
+ <a href="pg_hint_plan.html">pg_hint_plan</a>
</div>
<hr>
<li><a href="#examples">Hint descriptions</a></li>
<li><a href="#hint_syntax">Hint syntax</a></li>
<li><a href="#restrictions">Restrictions</a></li>
-<li><a href="#techniques">Techniques to hint on disired targets</a></li>
+<li><a href="#techniques">Techniques to hint on desired targets</a></li>
<li><a href="#errors">Errors of hints</a></li>
<li><a href="#func_limits">Functional limitations</a></li>
<li><a href="#requirement">Requirements</a></li>
<p>pg_hint_plan -- controls execution plan with hinting phrases in comment of special form.</p>
<h2 id="synopsis">Synopsis</h2>
-<p>PostgreSQL uses cost based optimizer, which utilizes data statistics, not static rules. The planner (optimizer) esitimates costs of each possible execution plans for a SQL statement then the execution plan with the lowest cost finally be executed. The planner does its best to select the best best execution plan, but not perfect, since it doesn't count some properties of the data, for example, correlation between columns.</p>
+<p>PostgreSQL uses cost based optimizer, which utilizes data statistics, not static rules. The planner (optimizer) estimates costs of each possible execution plan for a SQL statement, and then the execution plan with the lowest cost is eventually executed. The planner does its best to select the best execution plan, but it is not perfect, since it doesn't know all properties of the data, for example, correlation between columns.</p>
<p>pg_hint_plan makes it possible to tweak execution plans using so-called "hints", which are simple descriptions in the SQL comment of special form.</p>
<h2 id="description">Description</h2>
<h3 id="hint-rule">Basic Usage</h3>
-<p>pg_hint_plan reads hinting phrases in a comment of special form given with the target SQL statement. The special form is beginning by the character sequence "/*+" and ends with "*/". Hint phrases are consists of hint name and following parameters enclosed by parentheses and delimited by spaces. Each hinting phrases can be delimited by new lines for readability.</p>
+<p>pg_hint_plan reads hinting phrases in a comment of special form given with the target SQL statement. The special form is beginning by the character sequence "/*+" and ends with "*/". Hint phrases consist of hint name and following parameters enclosed by parentheses and delimited by spaces. Each hinting phrases can be delimited by new lines for readability.</p>
-<p>In the example below , hash join is selected as the joning method and scanning pgbench_accounts by sequential scan method.</p>
+<p>In the example below , hash join is selected as the join method and scanning pgbench_accounts by sequential scan method.</p>
<pre>
postgres=# /*+
-postgres*# <b><u>HashJoin(a b)</u></b>
-postgres*# <b><u>>SeqScan(a)</u></b>
+postgres*# <span class="strong">HashJoin(a b)</span>
+postgres*# <span class="strong">SeqScan(a)</span>
postgres*# */
postgres-# EXPLAIN SELECT *
postgres-# FROM pgbench_branches b
---------------------------------------------------------------------------------------
Sort (cost=31465.84..31715.84 rows=100000 width=197)
Sort Key: a.aid
- -> <b><u>Hash Join</u></b> (cost=1.02..4016.02 rows=100000 width=197)
+ -> <span class="strong">Hash Join</span> (cost=1.02..4016.02 rows=100000 width=197)
Hash Cond: (a.bid = b.bid)
- -> <b><u>Seq Scan on pgbench_accounts a</u></b> (cost=0.00..2640.00 rows=100000 width=97)
+ -> <span class="strong">Seq Scan on pgbench_accounts a</span> (cost=0.00..2640.00 rows=100000 width=97)
-> Hash (cost=1.01..1.01 rows=1 width=100)
-> Seq Scan on pgbench_branches b (cost=0.00..1.01 rows=1 width=100)
(7 rows)
postgres=# </pre>
-<h2 id="examples">Hint table</h2>
-<p> The above section mentioned that
-hints are placed in a comment with special form. This way of hinting
-is inconvenient for certain cases where queries cannot be edited. For
-such cases, hints can be placed in a special table named
-"hint_plan.hints". The table consists of the following columns.</p>
+<h2 id="examples">The hint table</h2>
+<p> Hints are described in a comment in a special form in the above
+section. This is inconvenient in the case where queries cannot be
+edited. In the case hints can be placed in a special table named
+"hint_plan.hints". This feature requires compute_query_id to be "on"
+or "auto". The table consists of the following columns.</p>
<table>
<thead>
<tr>
<p>The following example shows how to operate with the hint table.</p>
<pre>
-<b>postgres=#</b> INSERT INTO hint_plan.hints(norm_query_string, application_name, hints)
-<b>postgres-#</b> VALUES (
-<b>postgres(#</b> 'EXPLAIN (COSTS false) SELECT * FROM t1 WHERE t1.id = ?;',
-<b>postgres(#</b> '',
-<b>postgres(#</b> 'SeqScan(t1)'
-<b>postgres(#</b> );
+<span class="strong">postgres=#</span> INSERT INTO hint_plan.hints(norm_query_string, application_name, hints)
+<span class="strong">postgres-#</span> VALUES (
+<span class="strong">postgres(#</span> 'EXPLAIN (COSTS false) SELECT * FROM t1 WHERE t1.id = ?;',
+<span class="strong">postgres(#</span> '',
+<span class="strong">postgres(#</span> 'SeqScan(t1)'
+<span class="strong">postgres(#</span> );
INSERT 0 1
-<b>postgres=#</b> UPDATE hint_plan.hints
-<b>postgres-#</b> SET hints = 'IndexScan(t1)'
-<b>postgres-#</b> WHERE id = 1;
+<span class="strong">postgres=#</span> UPDATE hint_plan.hints
+<span class="strong">postgres-#</span> SET hints = 'IndexScan(t1)'
+<span class="strong">postgres-#</span> WHERE id = 1;
UPDATE 1
-<b>postgres=#</b> DELETE FROM hint_plan.hints
-<b>postgres-#</b> WHERE id = 1;
+<span class="strong">postgres=#</span> DELETE FROM hint_plan.hints
+<span class="strong">postgres-#</span> WHERE id = 1;
DELETE 1
-<b>postgres=#</b>
+<span class="strong">postgres=#</span>
</pre>
-<p>The hint tables is owned by the creator user and having the default previledges at the time of creation.
-during CREATE EXTENSION. Table hints are prioritized than comment hits.</p>
+<p>The hint table is owned by the creator user and having the default privileges at the time of creation during CREATE EXTENSION. Table hints are prioritized then comment hits.</p>
<h3 id="hint-group">The types of hints</h3>
<p>Hinting phrases are classified into six types based on what kind of object and how they can affect planning. Scaning methods, join methods, joining order, row number correction, parallel query and GUC setting. You will see the lists of hint phrases of each type in <a href="hint_list.html">Hint list</a>.</p>
<h4>Hints for scan methods </h4>
<p>Scan method hints enforce specific scanning method on the target table. pg_hint_plan recognizes the target table by alias names if any. They are 'SeqScan' , 'IndexScan' and so on in this kind of hint.</p>
-<p>Scan hints are effective on ordinary tables, inheritance tables, UNLOGGED tables, temporary tables and system catalogs. External(foreign) tables, table functions, VALUES clause, CTEs, views and subquiries are not affected.</p>
+<p>Scan hints are effective on ordinary tables, inheritance tables, UNLOGGED tables, temporary tables and system catalogs. External(foreign) tables, table functions, VALUES clause, CTEs, views and subqueries are not affected.</p>
<pre>
-<b>postgres=# /*+</b>
-<b>postgres*# SeqScan(t1)</b>
-<b>postgres*# IndexScan(t2 t2_pkey)</b>
-<b>postgres*# */</b>
-<b>postgres-#</b> SELECT * FROM table1 t1 JOIN table table2 t2 ON (t1.key = t2.key);
+<span class="strong">postgres=# /*+</span>
+<span class="strong">postgres*# SeqScan(t1)</span>
+<span class="strong">postgres*# IndexScan(t2 t2_pkey)</span>
+<span class="strong">postgres*# */</span>
+<span class="strong">postgres-#</span> SELECT * FROM table1 t1 JOIN table table2 t2 ON (t1.key = t2.key);
</pre>
<p>This can affect on joins only on ordinary tables, inheritance tables, UNLOGGED tables, temporary tables, external (foreign) tables, system catalogs, table functions, VALUES command results and CTEs are allowed to be in the parameter list. But joins on views and sub query are not affected.</p>
<h4>Hint for joining order</h4>
-<p> This hint "Leading" enforces the order of join on two or more tables. There are two ways of enforcing. One is enforcing specific order of joining but not restricting direction at each join level. Another enfoces join direction additionaly. Details are seen in the <a href="hint_list.html">hint list</a> table.</p>
+<p> This hint "Leading" enforces the order of join on two or more tables. There are two ways of enforcing. One is enforcing specific order of joining but not restricting direction at each join level. Another enforces join direction additionally. Details are seen in the <a href="hint_list.html">hint list</a> table.</p>
+
+<pre>
+<span class="strong">postgres=# /*+</span>
+<span class="strong">postgres*# NestLoop(t1 t2)</span>
+<span class="strong">postgres*# MergeJoin(t1 t2 t3)</span>
+<span class="strong">postgres*# Leading(t1 t2 t3)</span>
+<span class="strong">postgres*# */</span>
+<span class="strong">postgres-#</span> SELECT * FROM table1 t1
+<span class="strong">postgres-#</span> JOIN table table2 t2 ON (t1.key = t2.key)
+<span class="strong">postgres-#</span> JOIN table table3 t3 ON (t2.key = t3.key);
+</pre>
+<h4>Hint for restricting join behavior</h4>
+<p>This hint "Memoize" and "NoMemoize" allows and disallows a join to memoize the inner result. In the following example, The NoMemoize hint inhibits the topmost hash join from memoizing the result of table a.</p>
<pre>
-<b>postgres=# /*+</b>
-<b>postgres*# NestLoop(t1 t2)</b>
-<b>postgres*# MergeJoin(t1 t2 t3)</b>
-<b>postgres*# Leading(t1 t2 t3)</b>
-<b>postgres*# */</b>
-<b>postgres-#</b> SELECT * FROM table1 t1
-<b>postgres-#</b> JOIN table table2 t2 ON (t1.key = t2.key)
-<b>postgres-#</b> JOIN table table3 t3 ON (t2.key = t3.key);
+postgres=# /*+ NoMemoize(a b) */
+EXPLAIN SELECT * FROM a, b WHERE a.val = b.val;
+ QUERY PLAN
+--------------------------------------------------------------------
+ Hash Join (cost=270.00..1412.50 rows=100000 width=16)
+ Hash Cond: (b.val = a.val)
+ -> Seq Scan on b (cost=0.00..15.00 rows=1000 width=8)
+ -> Hash (cost=145.00..145.00 rows=10000 width=8)
+ -> Seq Scan on a (cost=0.00..145.00 rows=10000 width=8)
</pre>
+
+
<h4>Hint for row number correction</h4>
<p>This hint "Rows" corrects row number misestimation of joins that comes from restrictions of the planner. </p>
<pre>
-<b>postgres=# /*+ Rows(a b #10) */</b> SELECT... ; Sets rows of join result to 10
-<b>postgres=# /*+ Rows(a b +10) */</b> SELECT... ; Increments row number by 10
-<b>postgres=# /*+ Rows(a b -10) */</b> SELECT... ; Subtracts 10 from the row number.
-<b>postgres=# /*+ Rows(a b *10) */</b> SELECT... ; Makes the number 10 times larger.
+<span class="strong">postgres=# /*+ Rows(a b #10) */</span> SELECT... ; Sets rows of join result to 10
+<span class="strong">postgres=# /*+ Rows(a b +10) */</span> SELECT... ; Increments row number by 10
+<span class="strong">postgres=# /*+ Rows(a b -10) */</span> SELECT... ; Subtracts 10 from the row number.
+<span class="strong">postgres=# /*+ Rows(a b *10) */</span> SELECT... ; Makes the number 10 times larger.
</pre>
<h4>Hint for parallel plan</h4>
<p>This hint "Parallel" enforces parallel execution configuration on
-scans. The third parameter specifies the strength of enfocement. "soft" means that pg_hint_plan only changes max_parallel_worker_per_gather and leave all others to planner. "hard" changes other planner parameters so as to forcibly apply the number. This can affect on ordinary tables, inheritnce parents, unlogged
+scans. The third parameter specifies the strength of enforcement. "soft" means that pg_hint_plan only changes max_parallel_worker_per_gather and leave all others to planner. "hard" changes other planner parameters so as to forcibly apply the number. This can affect on ordinary tables, inheritance parents, unlogged
tables and system catalogues. External tables, table functions, values clause, CTEs, views and subqueries are not affected. Internal tables of a view can be specified by its real name/alias as the target object. The following example shows that the query is enforced differently on each table.</p>
<pre>
-<b>postgres=#</b> explain <b>/*+ Parallel(c1 3 hard) Parallel(c2 5 hard) */</b>
+<span class="strong">postgres=#</span> explain <span class="strong">/*+ Parallel(c1 3 hard) Parallel(c2 5 hard) */</span>
SELECT c2.a FROM c1 JOIN c2 ON (c1.a = c2.a);
QUERY PLAN
-------------------------------------------------------------------------------
Hash Join (cost=2.86..11406.38 rows=101 width=4)
Hash Cond: (c1.a = c2.a)
-> Gather (cost=0.00..7652.13 rows=1000101 width=4)
- <b><u>Workers Planned: 3</u></b>
- -> Parallel Seq Scan on <b><u>c1</u></b> (cost=0.00..7652.13 rows=322613 width=4)
+ <span class="strong">Workers Planned: 3</span>
+ -> Parallel Seq Scan on <span class="strong">c1</span> (cost=0.00..7652.13 rows=322613 width=4)
-> Hash (cost=1.59..1.59 rows=101 width=4)
-> Gather (cost=0.00..1.59 rows=101 width=4)
- <b><u>Workers Planned: 5</u></b>
- -> Parallel Seq Scan on <b><u>c2</u></b> (cost=0.00..1.59 rows=59 width=4)
+ <span class="strong">Workers Planned: 5</span>
+ -> Parallel Seq Scan on <span class="strong">c2</span> (cost=0.00..1.59 rows=59 width=4)
-<b>postgres=#</b> EXPLAIN <b>/*+ Parallel(tl 5 hard) */</b> SELECT sum(a) FROM tl;
+<span class="strong">postgres=#</span> EXPLAIN <span class="strong">/*+ Parallel(tl 5 hard) */</span> SELECT sum(a) FROM tl;
QUERY PLAN
-----------------------------------------------------------------------------------
- <b><u>Finalize Aggregate</u></b> (cost=693.02..693.03 rows=1 width=8)
+ <span class="strong">Finalize Aggregate</span> (cost=693.02..693.03 rows=1 width=8)
-> Gather (cost=693.00..693.01 rows=5 width=8)
- <b><u>Workers Planned: 5</u></b>
- -> <b><u>Partial Aggregate</u></b> (cost=693.00..693.01 rows=1 width=8)
+ <span class="strong">Workers Planned: 5</span>
+ -> <span class="strong">Partial Aggregate</span> (cost=693.00..693.01 rows=1 width=8)
-> Parallel Seq Scan on tl (cost=0.00..643.00 rows=20000 width=4)
</pre>
</dd>
<h4>GUC parameters temporarily setting</h4>
<p>'Set' hint changes GUC parameters just while planning. GUC parameter shown in <a href="http://www.postgresql.org/docs/current/static/runtime-config-query.html">Query Planning</a> can have the expected effects on planning unless any other hint conflicts with the planner method configuration parameters. The last one among hints on the same GUC parameter makes effect. <a href="#hint-GUC">GUC parameters for pg_hint_plan</a> are also settable by this hint but it won't work as your expectation. See <a href="#restrictions">Restrictions</a> for details.</p>
<pre>
-postgres=# <b>/*+ Set(random_page_cost 2.0) */</b>
+postgres=# <span class="strong">/*+ Set(random_page_cost 2.0) */</span>
postgres-# SELECT * FROM table1 t1 WHERE key = 'value';
...
</pre>
<table>
<thead>
<tr>
-<tr><th>Parameter name</th><th>discription</th><th>Default</th></tr>
+<tr><th>Parameter name</th><th>description</th><th>Default</th></tr>
</tr></thead>
<tbody>
<tr><td>pg_hint_plan.enable_hint</td>
- <td>True enbles pg_hint_plan.</td><td>on</td></tr>
+ <td>True enables pg_hint_plan.</td><td>on</td></tr>
<tr><td>pg_hint_plan.enable_hint_table</td>
<td>True enbles hinting by table. true or false.</td><td>off</td></tr>
<tr><td>pg_hint_plan.parse_messages</td>
<td>Specifies the log level of hint parse error. Valid values are error, warning, notice, info, log, debug<n>.</td><td>INFO</td></tr>
<tr><td>pg_hint_plan.debug_print</td>
- <td>Controls debug print and verbosity. Valid vaiues are off, on, detailed and verbose.</td><td>off</td></tr>
+ <td>Controls debug print and verbosity. Valid values are off, on, detailed and verbose.</td><td>off</td></tr>
<tr><td>pg_hint_plan.message_level</td>
- <td>Specifies message level of debug print. Valid values are error, warning, notice, info, log, debug<n>.</td><td>INFO</td></tr>
+ <td>Specifies message level of debug print. Valid values are error, warning, notice, info, log, debug<n>.</td><td>LOG</td></tr>
+<tr><td>pg_hint_plan.hints_anywhere</td>
+ <td>If it is on, pg_hint_plan reads hints ignoring SQL syntax. This allows to hints to be placed anywhere in a query but be cautious of false reads.</td><td>off</td></tr>
</tbody>
</table>
-
<h2 id="install">Installation</h2>
This section describes the installation steps.
<h3 id="build">building binary module</h3>
-<p>Simplly run "make" in the top of the source tree, then "make install" as appropriate user. The PATH environment variable should be set properly for the target PostgreSQL for this process. </p>
+<p>Simply run "make" in the top of the source tree, then "make install" as appropriate user. The PATH environment variable should be set properly for the target PostgreSQL for this process. </p>
<pre>
$ tar xzvf pg_hint_plan-1.x.x.tar.gz
$ cd pg_hint_plan-1.x.x
# make install
</pre>
-<h3 id="hint-load">Loding pg_hint_plan</h3>
-<p>Basically pg_hint_plan does not requires CREATE EXTENSION. Simplly loading it by LOAD command will activate it and of course you can load it globally by setting shared_preload_libraries in postgresql.conf. Or you might be interested in ALTER USER SET/ALTER DATABASE SET for automatic loading for specific sessions.
+<h3 id="hint-load">Loading pg_hint_plan</h3>
+<p>Basically pg_hint_plan does not requires CREATE EXTENSION. Simply loading it by LOAD command will activate it and of course you can load it globally by setting shared_preload_libraries in postgresql.conf. Or you might be interested in ALTER USER SET/ALTER DATABASE SET for automatic loading for specific sessions.
<pre>
postgres=# LOAD 'pg_hint_plan';
LOAD
</pre>
<h2 id="hint_syntax">Details in hinting</h2>
-<h3>Syntax and placement</h3>
+<dl>
+<dt>Syntax and placement</dt>
<dd>pg_hint_plan reads hints from only the first block comment and any characters except alphabets, digits, spaces, underscores, commas and parentheses stops parsing immediately. In the following example HashJoin(a b) and SeqScan(a) are parsed as hints but IndexScan(a) and MergeJoin(a b) are not.
<pre>
-<b>postgres=# /*+</b>
-<b>postgres*# HashJoin(a b)</b>
-<b>postgres*# SeqScan(a)</b>
-<b>postgres*# */</b>
-<b>postgres-# /*+ IndexScan(a) */</b>
-<b>postgres-#</b> EXPLAIN SELECT <b>/*+ MergeJoin(a b) */</b> *
-<b>postgres-#</b> FROM pgbench_branches b
-<b>postgres-#</b> JOIN pgbench_accounts a ON b.bid = a.bid
-<b>postgres-#</b> ORDER BY a.aid;
+<span class="strong">postgres=# /*+</span>
+<span class="strong">postgres*# HashJoin(a b)</span>
+<span class="strong">postgres*# SeqScan(a)</span>
+<span class="strong">postgres*# */</span>
+<span class="strong">postgres-# /*+ IndexScan(a) */</span>
+<span class="strong">postgres-#</span> EXPLAIN SELECT <span class="strong">/*+ MergeJoin(a b) */</span> *
+<span class="strong">postgres-#</span> FROM pgbench_branches b
+<span class="strong">postgres-#</span> JOIN pgbench_accounts a ON b.bid = a.bid
+<span class="strong">postgres-#</span> ORDER BY a.aid;
QUERY PLAN
---------------------------------------------------------------------------------------
Sort (cost=31465.84..31715.84 rows=100000 width=197)
Sort Key: a.aid
- -> <b><u>Hash Join</u></b> (cost=1.02..4016.02 rows=100000 width=197)
+ -> <span class="strong">Hash Join</span> (cost=1.02..4016.02 rows=100000 width=197)
Hash Cond: (a.bid = b.bid)
- -> <b><u>Seq Scan on pgbench_accounts a</u></b> (cost=0.00..2640.00 rows=100000 width=97)
+ -> <span class="strong">Seq Scan on pgbench_accounts a</span> (cost=0.00..2640.00 rows=100000 width=97)
-> Hash (cost=1.01..1.01 rows=1 width=100)
-> Seq Scan on pgbench_branches b (cost=0.00..1.01 rows=1 width=100)
(7 rows)
postgres=# </pre>
</dd>
-<h3>Using with PL/pgSQL</h3>
+<dt>Using with PL/pgSQL</dt>
<dd>pg_hint_plan works for queries in PL/pgSQL scripts with some restrictions.
<ul>
- <li>Hints affect only on the following kind of queires.
+ <li>Hints affect only on the following kind of queries.
<ul>
- <li>Queries that returns one row. (SELECT, INSERT, UPDATE and DELETE)</li>
- <li>Queries that returns multiple rows. (RETURN QUERY)</li>
+ <li>Queries that return one row. (SELECT, INSERT, UPDATE and DELETE)</li>
+ <li>Queries that return multiple rows. (RETURN QUERY)</li>
<li>Dynamic SQL statements. (EXECUTE)</li>
<li>Cursor open. (OPEN)</li>
- <li>Loop over result of a query (FOR)</li>
+ <li>Loop over results of a query (FOR)</li>
</ul>
<li>A hint comment have to be placed after the first word in a query
postgres$# id integer;
postgres$# cnt integer;
postgres$# BEGIN
-postgres$# SELECT <b><u>/*+ NoIndexScan(a) */</u></b> aid
+postgres$# SELECT <span class="strong">/*+ NoIndexScan(a) */</span> aid
postgres$# INTO id FROM pgbench_accounts a WHERE aid = $1;
-postgres$# SELECT <b><u>/*+ SeqScan(a) */</u></b> count(*)
+postgres$# SELECT <span class="strong">/*+ SeqScan(a) */</span> count(*)
postgres$# INTO cnt FROM pgbench_accounts a;
postgres$# RETURN id + cnt;
postgres$# END;
</pre>
</dd>
-<h3>Letter case in a hinted object</h3>
+<dt>Letter case in the object names</dt>
<dd>Unlike the way PostgreSQL handles object names, pg_hint_plan compares bare object names in hints against the database internal object names in case sensitive way. Therefore an object name TBL in a hint matches only "TBL" in database and does not match any unquoted names like TBL, tbl or Tbl.
</dd>
-<h3>Escaping special chacaters in object names</h3>
+<dt>Escaping special characters in object names</dt>
<dd>The objects as the hint parameter should be enclosed by double quotes if they includes parentheses, double quotes and white spaces. The escaping rule is the same as PostgreSQL.
</dd>
-<h3>Distinction between multiple occurances of a table</h3>
+<h3>Distinction between multiple occurences of a table</h3>
<dd>pg_hint_plan identifies the target object by using aliases if
-exists. This behavior is usable to point a specific occurance
-among multiple occurances of one table.
+exists. This behavior is usable to point a specific occurence
+among multiple occurences of one table.
<pre>
-<b>postgres=# /*+ HashJoin(t1 t1) */</b>
-<b>postgres-#</b> EXPLAIN SELECT * FROM s1.t1
-<b>postgres-#</b> JOIN public.t1 ON (s1.t1.id=public.t1.id);
+<span class="strong">postgres=# /*+ HashJoin(t1 t1) */</span>
+<span class="strong">postgres-#</span> EXPLAIN SELECT * FROM s1.t1
+<span class="strong">postgres-#</span> JOIN public.t1 ON (s1.t1.id=public.t1.id);
INFO: hint syntax error at or near "HashJoin(t1 t1)"
-<b><u>DETAIL: Relation name "t1" is ambiguous.</u></b>
+<span class="strong">DETAIL: Relation name "t1" is ambiguous.</span>
...
-<b>postgres=# /*+ HashJoin(pt st) */</b>
-<b>postgres-#</b> EXPLAIN SELECT * FROM s1.t1 st
-<b>postgres-#</b> JOIN public.t1 pt ON (st.id=pt.id);
+<span class="strong">postgres=# /*+ HashJoin(pt st) */</span>
+<span class="strong">postgres-#</span> EXPLAIN SELECT * FROM s1.t1 st
+<span class="strong">postgres-#</span> JOIN public.t1 pt ON (st.id=pt.id);
QUERY PLAN
---------------------------------------------------------------------
<span class="strong">Hash Join</span> (cost=64.00..1112.00 rows=28800 width=8)
</pre>
</dd>
-<h3>Underlying tables of views or rules</h3>
+<dt>Underlying tables of views or rules</dt>
<dd>Hints are not applicable on views itself, but they can affect the
queries within if the object names match the object names in the
expanded query on the view. Assigning aliases to the tables in a view
enables them to be manipulated from outside the view.
<pre>
-<b>postgres=#</b> CREATE VIEW v1 AS SELECT * FROM <b><u>t2</u></b>;
-<b>postgres=#</b> EXPLAIN <b>/*+ HashJoin(t1 v1) */</b>
+<span class="strong">postgres=#</span> CREATE VIEW v1 AS SELECT * FROM <span class="strong">t2</span>;
+<span class="strong">postgres=#</span> EXPLAIN <span class="strong">/*+ HashJoin(t1 v1) */</span>
SELECT * FROM t1 JOIN v1 ON (c1.a = v1.a);
QUERY PLAN
------------------------------------------------------------------
</pre>
</dd>
-<h3>Inheritance tables</h3>
+<dt>Inheritance tables</dt>
<dd>Hints can point only the parent of an inheritance tables and the
hint affect all the inheritance. Hints simultaneously point directly
to children are not in effect.
</dd>
-<h3>Hinting on multistatements</h3>
+<dt>Hinting on multistatements</dt>
<dd>One multistatement can have exactly one hint comment and the hints affects all of the individual statement in the multistatement. Notice that the seemingly multistatement on the interactive interface of psql is internally a sequence of single statements so hints affects only on the statement just following.</dd>
-<h3>VALUES expressions</h3>
+<dt>VALUES expressions</dt>
<dd>VALUES expressions in FROM clause are named as *VALUES* internally
so it is hintable if it is the only VALUES in a query. Two or more
VALUES expressions in a query seems distinguishable looking its
explain result. But in reality it is mere a cosmetic and they are not
-distinguisable.
+distinguishable.
<pre>
-<b>postgres=# /*+ MergeJoin(*VALUES*_1 *VALUES*) */</b>
+<span class="strong">postgres=#</span> <span class="strong">/*+ MergeJoin(*VALUES*_1 *VALUES*) */</span>
EXPLAIN SELECT * FROM (VALUES (1, 1), (2, 2)) v (a, b)
JOIN (VALUES (1, 5), (2, 8), (3, 4)) w (a, c) ON v.a = w.a;
INFO: pg_hint_plan: hint syntax error at or near "MergeJoin(*VALUES*_1 *VALUES*) "
-<b><u>DETAIL: Relation name "*VALUES*" is ambiguous.</u></b>
+<span class="strong">DETAIL: Relation name "*VALUES*" is ambiguous.</span>
QUERY PLAN
-------------------------------------------------------------------------
Hash Join (cost=0.05..0.12 rows=2 width=16)
<p>Subqueries in the following context occasionally can be hinted using the
name "ANY_subquery".</p>
<pre>
-IN (SELECT ... {<b>LIMIT | OFFSET</b> ...} ...)
-= ANY (SELECT ... {<b>LIMIT | OFFSET</b> ...} ...)
-= SOME (SELECT ... {<b>LIMIT | OFFSET</b> ...} ...)
+IN (SELECT ... {<span class="strong">LIMIT | OFFSET</span> ...} ...)
+= ANY (SELECT ... {<span class="strong">LIMIT | OFFSET</span> ...} ...)
+= SOME (SELECT ... {<span class="strong">LIMIT | OFFSET</span> ...} ...)
</pre>
<p>For these syntaxes, planner internally assigns the name to the subquery when planning joins on tables including it, so join hints are applicable on such joins using the implicit name as the following.</p>
<pre>
-<b>postgres=# /*+HashJoin(a1 ANY_subquery)*/</b>
-<b>postgres=#</b> EXPLAIN SELECT *
-<b>postgres=#</b> FROM pgbench_accounts a1
-<b>postgres=#</b> WHERE aid IN (SELECT bid FROM pgbench_accounts a2 LIMIT 10);
+<span class="strong">postgres=# /*+HashJoin(a1 ANY_subquery)*/</span>
+<span class="strong">postgres=#</span> EXPLAIN SELECT *
+<span class="strong">postgres=#</span> FROM pgbench_accounts a1
+<span class="strong">postgres=#</span> WHERE aid IN (SELECT bid FROM pgbench_accounts a2 LIMIT 10);
QUERY PLAN
---------------------------------------------------------------------------------------------
- <b><u>Hash Semi Join</u></b> (cost=0.49..2903.00 rows=1 width=97)
+ <span class="strong">Hash Semi Join</span> (cost=0.49..2903.00 rows=1 width=97)
Hash Cond: (a1.aid = a2.bid)
-> Seq Scan on pgbench_accounts a1 (cost=0.00..2640.00 rows=100000 width=97)
-> Hash (cost=0.36..0.36 rows=10 width=4)
</pre>
</dd>
-<h3>Using IndexOnlyScan hint</h3>
-<dd>An additional hint to enforce an index that can perform index only scan is required for IndexOnlyScan hint to work when any index that can not perform index only scan is also defined on the same table. </dd>
+<dt>Using IndexOnlyScan hint</dt>
+<dd>Index scan may unexpectedly performed on another index when the index specified in IndexOnlyScan hint cannot perform index only scan.</dd>
-<h3>Behavior of NoIndexScan</h3>
-<dd>NoIndexScan hint involes NoIndexOnlyScan.</dd>
+<dt>Behavior of NoIndexScan</dt>
+<dd>NoIndexScan hint involves NoIndexOnlyScan.</dd>
-<h3>Parallel hint and UNION</h3>
+<dt>Parallel hint and UNION</dt>
<dd>A UNION can run in parallel only when all underlying subqueries
are parallel-safe. Conversely enforcing parallel on any of
the subqueries let a parallel-executable UNION run in
-parallel. Meanwhile, a parallel hint with zero workers hinhibits a scan
-from executed in parallel.</dd>
+parallel. Meanwhile, a parallel hint with zero workers inhibits a scan
+from being executed in parallel.</dd>
-<h3>Setting pg_hint_plan parameters by Set hints</h3>
-<dd><p>pg_hint_plan paramters change the behavior of itself so some parameters doesn't work as expected.</p>
+<dt>Setting pg_hint_plan parameters by Set hints</dt>
+<dd><p>pg_hint_plan parameters change the behavior of itself so some parameters doesn't work as expected.</p>
<ul>
<li>Hints to change enable_hint, enable_hint_tables are ignored even though they are reported as "used hints" in debug logs.</li>
<li>Setting debug_print and message_level works from midst of the processing of the target query.</li>
</ul>
</dd>
-
+</dl>
<h2 id="errors">Errors</h2>
<p>pg_hint_plan stops parsing on any error and uses hints already parsed on the most cases. Followings are the typical errors.</p>
-<h3>Syntax errors </h3>
-<p>Any syntactical errors or wrong hint names are reported as an syntax error. These errors are reported in the server log with the message level which specified by pg_hint_plan.message_level if pg_hint_plan.debug_print is on and aboves.</p>
-<h3>Object misspecifications</h3>
-<p>Object misspecifications results silent ingorance of the hints. This kind of error is reported as "not used hints" in the server log by the same condtion to syntax errors.</p>
+<dl>
+<dt>Syntax errors </dt>
+<dd>Any syntactical errors or wrong hint names are reported as an syntax error. These errors are reported in the server log with the message level which specified by pg_hint_plan.message_level if pg_hint_plan.debug_print is on and aboves.</dd>
+<dt>Object misspecifications</dt>
+<dd>Object misspecifications results silent ingnorance of the hints. This kind of error is reported as "not used hints" in the server log by the same condition to syntax errors.</dd>
-<h3>Redundant or conflicting hints</h3>
-<p>The last hint will be active when redundant hints or hints conflicting with each other. This kind of error is reported as "duplication hints" in the server log by the same condition to syntax errors.</p>
+<dt>Redundant or conflicting hints</dt>
+<dd>The last hint will be active when redundant hints or hints conflicting with each other. This kind of error is reported as "duplication hints" in the server log by the same condition to syntax errors.</dd>
-<h3>Nested comments</h3>
-<p>Hint comment cannot include another block comment within. If pg_hint_plan finds it, differently from other erros, it stops parsing and abandans all hints already parsed. This kind of error is reported in the same manner as other errors. </p>
+<dt>Nested comments</dt>
+<dd>Hint comment cannot include another block comment within. If pg_hint_plan finds it, differently from other errors, it stops parsing and abandons all hints already parsed. This kind of error is reported in the same manner as other errors. </dd>
<h2 id="func_limits">Functional limitations</h2>
-<h3>Influences of some of planner GUC parameters</h3>
-<p>The planner does not try to consider joining order for FROM clause entries more than from_collapse_limit. pg_hint_plan cannot affect joining order as expected for the case.</p>
+<dt>Influences of some of planner GUC parameters</dt>
+<dd>The planner does not try to consider joining order for FROM clause entries more than from_collapse_limit. pg_hint_plan cannot affect joining order as expected for the case.</dd>
-<h3>Cases that pg_hint_plan essentially does not work </h3>
-<p>By the nature of pg_hint_plan, it cannot affect some cases that out of scope of the planner like following.</p>
+<dt>Hints trying to enforce unexecutable plans </dt>
+<dd>Planner chooses any executable plans when the enforced plan cannot be executed.
<ul>
<li>FULL OUTER JOIN to use nested loop</li>
<li>To use indexes that does not have columns used in quals</li>
<li>To do TID scans for queries without ctid conditions</li>
</ul>
+</dd>
-<h3>Queries in ECPG</h3>
-<p>ECPG removes comments in queries written as embedded SQLs so hints cannot be passed form those queries. The only exception is that EXECUTE command passes given string unmodifed. Please consider hint tables for this case.</p>
+<dt>Queries in ECPG</dt>
+<dd>ECPG removes comments in queries written as embedded SQLs so hints cannot be passed form those queries. The only exception is that EXECUTE command passes given string unmodified. Please consider hint tables in the case.</dd>
-<h3>Work with pg_stat_statements</h3>
-<p>pg_stat_statements generates a query id ignoring comments. As the result the identical queires with different hints are summerized as the same query.</p>
+<dt>Work with pg_stat_statements</dt>
+<dd>pg_stat_statements generates a query id ignoring comments. As the result the identical queires with different hints are summarized as the same query.</dd>
<h2 id="requirement">Requirements</h2>
-pg_hint_plan 1.3 requires PostgreSQL 10.
+pg_hint_plan14 1.6 requires PostgreSQL 14.
<dl>
<dt>PostgreSQL versions tested</dt>
- <dd>Version 10</dd>
+ <dd>Version 14</dd>
<dt>OS versions tested</dt>
- <dd>CentOS 7.4</dd>
+ <dd>CentOS 8.5</dd>
</dl>
<h2 id="seealso">See also</h2>
<a href="pg_hint_plan.html">pg_hint_plan</a>
</div>
-<p class="footer">Copyright (c) 2012-2018, NIPPON TELEGRAPH AND TELEPHONE CORPORATION</p>
+<p class="footer">Copyright (c) 2012-2022, NIPPON TELEGRAPH AND TELEPHONE CORPORATION</p>
</body>
</html>
OSDN Git Service
