1EXPLAIN(7) PostgreSQL 16.1 Documentation EXPLAIN(7)
2
6 EXPLAIN - show the execution plan of a statement
7
9 EXPLAIN [ ( option [, ...] ) ] statement
10 EXPLAIN [ ANALYZE ] [ VERBOSE ] statement
11 where option can be one of:
13 ANALYZE [ boolean ]
15 VERBOSE [ boolean ]
16 COSTS [ boolean ]
17 SETTINGS [ boolean ]
18 GENERIC_PLAN [ boolean ]
19 BUFFERS [ boolean ]
20 WAL [ boolean ]
21 TIMING [ boolean ]
22 SUMMARY [ boolean ]
23 FORMAT { TEXT | XML | JSON | YAML }
24
26 This command displays the execution plan that the PostgreSQL planner
27 generates for the supplied statement. The execution plan shows how the
28 table(s) referenced by the statement will be scanned — by plain
29 sequential scan, index scan, etc. — and if multiple tables are
30 referenced, what join algorithms will be used to bring together the
31 required rows from each input table.
32 The most critical part of the display is the estimated statement
34 execution cost, which is the planner's guess at how long it will take
35 to run the statement (measured in cost units that are arbitrary, but
36 conventionally mean disk page fetches). Actually two numbers are shown:
37 the start-up cost before the first row can be returned, and the total
38 cost to return all the rows. For most queries the total cost is what
39 matters, but in contexts such as a subquery in EXISTS, the planner will
40 choose the smallest start-up cost instead of the smallest total cost
41 (since the executor will stop after getting one row, anyway). Also, if
42 you limit the number of rows to return with a LIMIT clause, the planner
43 makes an appropriate interpolation between the endpoint costs to
44 estimate which plan is really the cheapest.
45 The ANALYZE option causes the statement to be actually executed, not
47 only planned. Then actual run time statistics are added to the display,
48 including the total elapsed time expended within each plan node (in
49 milliseconds) and the total number of rows it actually returned. This
50 is useful for seeing whether the planner's estimates are close to
51 reality.
52 Important
54 Keep in mind that the statement is actually executed when the
55 ANALYZE option is used. Although EXPLAIN will discard any output
56 that a SELECT would return, other side effects of the statement
57 will happen as usual. If you wish to use EXPLAIN ANALYZE on an
58 INSERT, UPDATE, DELETE, MERGE, CREATE TABLE AS, or EXECUTE
59 statement without letting the command affect your data, use this
60 approach:
61 BEGIN;
63 EXPLAIN ANALYZE ...;
64 ROLLBACK;
65 Only the ANALYZE and VERBOSE options can be specified, and only in that
67 order, without surrounding the option list in parentheses. Prior to
68 PostgreSQL 9.0, the unparenthesized syntax was the only one supported.
69 It is expected that all new options will be supported only in the
70 parenthesized syntax.
71
73 ANALYZE
74 Carry out the command and show actual run times and other
75 statistics. This parameter defaults to FALSE.
76 VERBOSE
78 Display additional information regarding the plan. Specifically,
79 include the output column list for each node in the plan tree,
80 schema-qualify table and function names, always label variables in
81 expressions with their range table alias, and always print the name
82 of each trigger for which statistics are displayed. The query
83 identifier will also be displayed if one has been computed, see
84 compute_query_id for more details. This parameter defaults to
85 FALSE.
86 COSTS
88 Include information on the estimated startup and total cost of each
89 plan node, as well as the estimated number of rows and the
90 estimated width of each row. This parameter defaults to TRUE.
91 SETTINGS
93 Include information on configuration parameters. Specifically,
94 include options affecting query planning with value different from
95 the built-in default value. This parameter defaults to FALSE.
96 GENERIC_PLAN
98 Allow the statement to contain parameter placeholders like $1, and
99 generate a generic plan that does not depend on the values of those
100 parameters. See PREPARE for details about generic plans and the
101 types of statement that support parameters. This parameter cannot
102 be used together with ANALYZE. It defaults to FALSE.
103 BUFFERS
105 Include information on buffer usage. Specifically, include the
106 number of shared blocks hit, read, dirtied, and written, the number
107 of local blocks hit, read, dirtied, and written, the number of temp
108 blocks read and written, and the time spent reading and writing
109 data file blocks and temporary file blocks (in milliseconds) if
110 track_io_timing is enabled. A hit means that a read was avoided
111 because the block was found already in cache when needed. Shared
112 blocks contain data from regular tables and indexes; local blocks
113 contain data from temporary tables and indexes; while temporary
114 blocks contain short-term working data used in sorts, hashes,
115 Materialize plan nodes, and similar cases. The number of blocks
116 dirtied indicates the number of previously unmodified blocks that
117 were changed by this query; while the number of blocks written
118 indicates the number of previously-dirtied blocks evicted from
119 cache by this backend during query processing. The number of blocks
120 shown for an upper-level node includes those used by all its child
121 nodes. In text format, only non-zero values are printed. This
122 parameter defaults to FALSE.
123 WAL
125 Include information on WAL record generation. Specifically, include
126 the number of records, number of full page images (fpi) and the
127 amount of WAL generated in bytes. In text format, only non-zero
128 values are printed. This parameter may only be used when ANALYZE is
129 also enabled. It defaults to FALSE.
130 TIMING
132 Include actual startup time and time spent in each node in the
133 output. The overhead of repeatedly reading the system clock can
134 slow down the query significantly on some systems, so it may be
135 useful to set this parameter to FALSE when only actual row counts,
136 and not exact times, are needed. Run time of the entire statement
137 is always measured, even when node-level timing is turned off with
138 this option. This parameter may only be used when ANALYZE is also
139 enabled. It defaults to TRUE.
140 SUMMARY
142 Include summary information (e.g., totaled timing information)
143 after the query plan. Summary information is included by default
144 when ANALYZE is used but otherwise is not included by default, but
145 can be enabled using this option. Planning time in EXPLAIN EXECUTE
146 includes the time required to fetch the plan from the cache and the
147 time required for re-planning, if necessary.
148 FORMAT
150 Specify the output format, which can be TEXT, XML, JSON, or YAML.
151 Non-text output contains the same information as the text output
152 format, but is easier for programs to parse. This parameter
153 defaults to TEXT.
154 boolean
156 Specifies whether the selected option should be turned on or off.
157 You can write TRUE, ON, or 1 to enable the option, and FALSE, OFF,
158 or 0 to disable it. The boolean value can also be omitted, in which
159 case TRUE is assumed.
160 statement
162 Any SELECT, INSERT, UPDATE, DELETE, MERGE, VALUES, EXECUTE,
163 DECLARE, CREATE TABLE AS, or CREATE MATERIALIZED VIEW AS statement,
164 whose execution plan you wish to see.
165
167 The command's result is a textual description of the plan selected for
168 the statement, optionally annotated with execution statistics.
169 Section 14.1 describes the information provided.
170
172 In order to allow the PostgreSQL query planner to make reasonably
173 informed decisions when optimizing queries, the pg_statistic data
174 should be up-to-date for all tables used in the query. Normally the
175 autovacuum daemon will take care of that automatically. But if a table
176 has recently had substantial changes in its contents, you might need to
177 do a manual ANALYZE rather than wait for autovacuum to catch up with
178 the changes.
179 In order to measure the run-time cost of each node in the execution
181 plan, the current implementation of EXPLAIN ANALYZE adds profiling
182 overhead to query execution. As a result, running EXPLAIN ANALYZE on a
183 query can sometimes take significantly longer than executing the query
184 normally. The amount of overhead depends on the nature of the query, as
185 well as the platform being used. The worst case occurs for plan nodes
186 that in themselves require very little time per execution, and on
187 machines that have relatively slow operating system calls for obtaining
188 the time of day.
189
191 To show the plan for a simple query on a table with a single integer
192 column and 10000 rows:
193 EXPLAIN SELECT * FROM foo;
195 QUERY PLAN
197 ---------------------------------------------------------
198 Seq Scan on foo (cost=0.00..155.00 rows=10000 width=4)
199 (1 row)
200 Here is the same query, with JSON output formatting:
202 EXPLAIN (FORMAT JSON) SELECT * FROM foo;
204 QUERY PLAN
205 --------------------------------
206 [ +
207 { +
208 "Plan": { +
209 "Node Type": "Seq Scan",+
210 "Relation Name": "foo", +
211 "Alias": "foo", +
212 "Startup Cost": 0.00, +
213 "Total Cost": 155.00, +
214 "Plan Rows": 10000, +
215 "Plan Width": 4 +
216 } +
217 } +
218 ]
219 (1 row)
220 If there is an index and we use a query with an indexable WHERE
222 condition, EXPLAIN might show a different plan:
223 EXPLAIN SELECT * FROM foo WHERE i = 4;
225 QUERY PLAN
227 --------------------------------------------------------------
228 Index Scan using fi on foo (cost=0.00..5.98 rows=1 width=4)
229 Index Cond: (i = 4)
230 (2 rows)
231 Here is the same query, but in YAML format:
233 EXPLAIN (FORMAT YAML) SELECT * FROM foo WHERE i='4';
235 QUERY PLAN
236 -------------------------------
237 - Plan: +
238 Node Type: "Index Scan" +
239 Scan Direction: "Forward"+
240 Index Name: "fi" +
241 Relation Name: "foo" +
242 Alias: "foo" +
243 Startup Cost: 0.00 +
244 Total Cost: 5.98 +
245 Plan Rows: 1 +
246 Plan Width: 4 +
247 Index Cond: "(i = 4)"
248 (1 row)
249 XML format is left as an exercise for the reader.
251 Here is the same plan with cost estimates suppressed:
253 EXPLAIN (COSTS FALSE) SELECT * FROM foo WHERE i = 4;
255 QUERY PLAN
257 ----------------------------
258 Index Scan using fi on foo
259 Index Cond: (i = 4)
260 (2 rows)
261 Here is an example of a query plan for a query using an aggregate
263 function:
264 EXPLAIN SELECT sum(i) FROM foo WHERE i < 10;
266 QUERY PLAN
268 ---------------------------------------------------------------------
269 Aggregate (cost=23.93..23.93 rows=1 width=4)
270 -> Index Scan using fi on foo (cost=0.00..23.92 rows=6 width=4)
271 Index Cond: (i < 10)
272 (3 rows)
273 Here is an example of using EXPLAIN EXECUTE to display the execution
275 plan for a prepared query:
276 PREPARE query(int, int) AS SELECT sum(bar) FROM test
278 WHERE id > $1 AND id < $2
279 GROUP BY foo;
280 EXPLAIN ANALYZE EXECUTE query(100, 200);
282 QUERY PLAN
284 -------------------------------------------------------------------------------------------------------------------------
285 HashAggregate (cost=10.77..10.87 rows=10 width=12) (actual time=0.043..0.044 rows=10 loops=1)
286 Group Key: foo
287 Batches: 1 Memory Usage: 24kB
288 -> Index Scan using test_pkey on test (cost=0.29..10.27 rows=99 width=8) (actual time=0.009..0.025 rows=99 loops=1)
289 Index Cond: ((id > 100) AND (id < 200))
290 Planning Time: 0.244 ms
291 Execution Time: 0.073 ms
292 (7 rows)
293 Of course, the specific numbers shown here depend on the actual
295 contents of the tables involved. Also note that the numbers, and even
296 the selected query strategy, might vary between PostgreSQL releases due
297 to planner improvements. In addition, the ANALYZE command uses random
298 sampling to estimate data statistics; therefore, it is possible for
299 cost estimates to change after a fresh run of ANALYZE, even if the
300 actual distribution of data in the table has not changed.
301 Notice that the previous example showed a “custom” plan for the
303 specific parameter values given in EXECUTE. We might also wish to see
304 the generic plan for a parameterized query, which can be done with
305 GENERIC_PLAN:
306 EXPLAIN (GENERIC_PLAN)
308 SELECT sum(bar) FROM test
309 WHERE id > $1 AND id < $2
310 GROUP BY foo;
311 QUERY PLAN
313 -------------------------------------------------------------------------------
314 HashAggregate (cost=26.79..26.89 rows=10 width=12)
315 Group Key: foo
316 -> Index Scan using test_pkey on test (cost=0.29..24.29 rows=500 width=8)
317 Index Cond: ((id > $1) AND (id < $2))
318 (4 rows)
319 In this case the parser correctly inferred that $1 and $2 should have
321 the same data type as id, so the lack of parameter type information
322 from PREPARE was not a problem. In other cases it might be necessary to
323 explicitly specify types for the parameter symbols, which can be done
324 by casting them, for example:
325 EXPLAIN (GENERIC_PLAN)
327 SELECT sum(bar) FROM test
328 WHERE id > $1::integer AND id < $2::integer
329 GROUP BY foo;
330
333 There is no EXPLAIN statement defined in the SQL standard.
334
336 ANALYZE(7)
337PostgreSQL 16.1 2023 EXPLAIN(7)