1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
|
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>CREATE INDEX</title><link rel="stylesheet" type="text/css" href="stylesheet.css" /><link rev="made" href="pgsql-docs@lists.postgresql.org" /><meta name="generator" content="DocBook XSL Stylesheets Vsnapshot" /><link rel="prev" href="sql-creategroup.html" title="CREATE GROUP" /><link rel="next" href="sql-createlanguage.html" title="CREATE LANGUAGE" /></head><body id="docContent" class="container-fluid col-10"><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">CREATE INDEX</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="sql-creategroup.html" title="CREATE GROUP">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="sql-commands.html" title="SQL Commands">Up</a></td><th width="60%" align="center">SQL Commands</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 15.6 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="sql-createlanguage.html" title="CREATE LANGUAGE">Next</a></td></tr></table><hr /></div><div class="refentry" id="SQL-CREATEINDEX"><div class="titlepage"></div><a id="id-1.9.3.69.1" class="indexterm"></a><div class="refnamediv"><h2><span class="refentrytitle">CREATE INDEX</span></h2><p>CREATE INDEX — define a new index</p></div><div class="refsynopsisdiv"><h2>Synopsis</h2><pre class="synopsis">
CREATE [ UNIQUE ] INDEX [ CONCURRENTLY ] [ [ IF NOT EXISTS ] <em class="replaceable"><code>name</code></em> ] ON [ ONLY ] <em class="replaceable"><code>table_name</code></em> [ USING <em class="replaceable"><code>method</code></em> ]
( { <em class="replaceable"><code>column_name</code></em> | ( <em class="replaceable"><code>expression</code></em> ) } [ COLLATE <em class="replaceable"><code>collation</code></em> ] [ <em class="replaceable"><code>opclass</code></em> [ ( <em class="replaceable"><code>opclass_parameter</code></em> = <em class="replaceable"><code>value</code></em> [, ... ] ) ] ] [ ASC | DESC ] [ NULLS { FIRST | LAST } ] [, ...] )
[ INCLUDE ( <em class="replaceable"><code>column_name</code></em> [, ...] ) ]
[ NULLS [ NOT ] DISTINCT ]
[ WITH ( <em class="replaceable"><code>storage_parameter</code></em> [= <em class="replaceable"><code>value</code></em>] [, ... ] ) ]
[ TABLESPACE <em class="replaceable"><code>tablespace_name</code></em> ]
[ WHERE <em class="replaceable"><code>predicate</code></em> ]
</pre></div><div class="refsect1" id="id-1.9.3.69.5"><h2>Description</h2><p>
<code class="command">CREATE INDEX</code> constructs an index on the specified column(s)
of the specified relation, which can be a table or a materialized view.
Indexes are primarily used to enhance database performance (though
inappropriate use can result in slower performance).
</p><p>
The key field(s) for the index are specified as column names,
or alternatively as expressions written in parentheses.
Multiple fields can be specified if the index method supports
multicolumn indexes.
</p><p>
An index field can be an expression computed from the values of
one or more columns of the table row. This feature can be used
to obtain fast access to data based on some transformation of
the basic data. For example, an index computed on
<code class="literal">upper(col)</code> would allow the clause
<code class="literal">WHERE upper(col) = 'JIM'</code> to use an index.
</p><p>
<span class="productname">PostgreSQL</span> provides the index methods
B-tree, hash, GiST, SP-GiST, GIN, and BRIN. Users can also define their own
index methods, but that is fairly complicated.
</p><p>
When the <code class="literal">WHERE</code> clause is present, a
<em class="firstterm">partial index</em> is created.
A partial index is an index that contains entries for only a portion of
a table, usually a portion that is more useful for indexing than the
rest of the table. For example, if you have a table that contains both
billed and unbilled orders where the unbilled orders take up a small
fraction of the total table and yet that is an often used section, you
can improve performance by creating an index on just that portion.
Another possible application is to use <code class="literal">WHERE</code> with
<code class="literal">UNIQUE</code> to enforce uniqueness over a subset of a
table. See <a class="xref" href="indexes-partial.html" title="11.8. Partial Indexes">Section 11.8</a> for more discussion.
</p><p>
The expression used in the <code class="literal">WHERE</code> clause can refer
only to columns of the underlying table, but it can use all columns,
not just the ones being indexed. Presently, subqueries and
aggregate expressions are also forbidden in <code class="literal">WHERE</code>.
The same restrictions apply to index fields that are expressions.
</p><p>
All functions and operators used in an index definition must be
<span class="quote">“<span class="quote">immutable</span>”</span>, that is, their results must depend only on
their arguments and never on any outside influence (such as
the contents of another table or the current time). This restriction
ensures that the behavior of the index is well-defined. To use a
user-defined function in an index expression or <code class="literal">WHERE</code>
clause, remember to mark the function immutable when you create it.
</p></div><div class="refsect1" id="id-1.9.3.69.6"><h2>Parameters</h2><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="literal">UNIQUE</code></span></dt><dd><p>
Causes the system to check for
duplicate values in the table when the index is created (if data
already exist) and each time data is added. Attempts to
insert or update data which would result in duplicate entries
will generate an error.
</p><p>
Additional restrictions apply when unique indexes are applied to
partitioned tables; see <a class="xref" href="sql-createtable.html" title="CREATE TABLE"><span class="refentrytitle">CREATE TABLE</span></a>.
</p></dd><dt><span class="term"><code class="literal">CONCURRENTLY</code></span></dt><dd><p>
When this option is used, <span class="productname">PostgreSQL</span> will build the
index without taking any locks that prevent concurrent inserts,
updates, or deletes on the table; whereas a standard index build
locks out writes (but not reads) on the table until it's done.
There are several caveats to be aware of when using this option
— see <a class="xref" href="sql-createindex.html#SQL-CREATEINDEX-CONCURRENTLY" title="Building Indexes Concurrently">Building Indexes Concurrently</a> below.
</p><p>
For temporary tables, <code class="command">CREATE INDEX</code> is always
non-concurrent, as no other session can access them, and
non-concurrent index creation is cheaper.
</p></dd><dt><span class="term"><code class="literal">IF NOT EXISTS</code></span></dt><dd><p>
Do not throw an error if a relation with the same name already exists.
A notice is issued in this case. Note that there is no guarantee that
the existing index is anything like the one that would have been created.
Index name is required when <code class="literal">IF NOT EXISTS</code> is specified.
</p></dd><dt><span class="term"><code class="literal">INCLUDE</code></span></dt><dd><p>
The optional <code class="literal">INCLUDE</code> clause specifies a
list of columns which will be included in the index
as <em class="firstterm">non-key</em> columns. A non-key column cannot
be used in an index scan search qualification, and it is disregarded
for purposes of any uniqueness or exclusion constraint enforced by
the index. However, an index-only scan can return the contents of
non-key columns without having to visit the index's table, since
they are available directly from the index entry. Thus, addition of
non-key columns allows index-only scans to be used for queries that
otherwise could not use them.
</p><p>
It's wise to be conservative about adding non-key columns to an
index, especially wide columns. If an index tuple exceeds the
maximum size allowed for the index type, data insertion will fail.
In any case, non-key columns duplicate data from the index's table
and bloat the size of the index, thus potentially slowing searches.
Furthermore, B-tree deduplication is never used with indexes
that have a non-key column.
</p><p>
Columns listed in the <code class="literal">INCLUDE</code> clause don't need
appropriate operator classes; the clause can include
columns whose data types don't have operator classes defined for
a given access method.
</p><p>
Expressions are not supported as included columns since they cannot be
used in index-only scans.
</p><p>
Currently, the B-tree, GiST and SP-GiST index access methods support
this feature. In these indexes, the values of columns listed
in the <code class="literal">INCLUDE</code> clause are included in leaf tuples
which correspond to heap tuples, but are not included in upper-level
index entries used for tree navigation.
</p></dd><dt><span class="term"><em class="replaceable"><code>name</code></em></span></dt><dd><p>
The name of the index to be created. No schema name can be included
here; the index is always created in the same schema as its parent
table. The name of the index must be distinct from the name of any
other relation (table, sequence, index, view, materialized view, or
foreign table) in that schema.
If the name is omitted, <span class="productname">PostgreSQL</span> chooses a
suitable name based on the parent table's name and the indexed column
name(s).
</p></dd><dt><span class="term"><code class="literal">ONLY</code></span></dt><dd><p>
Indicates not to recurse creating indexes on partitions, if the
table is partitioned. The default is to recurse.
</p></dd><dt><span class="term"><em class="replaceable"><code>table_name</code></em></span></dt><dd><p>
The name (possibly schema-qualified) of the table to be indexed.
</p></dd><dt><span class="term"><em class="replaceable"><code>method</code></em></span></dt><dd><p>
The name of the index method to be used. Choices are
<code class="literal">btree</code>, <code class="literal">hash</code>,
<code class="literal">gist</code>, <code class="literal">spgist</code>, <code class="literal">gin</code>,
<code class="literal">brin</code>, or user-installed access methods like
<a class="link" href="bloom.html" title="F.7. bloom">bloom</a>.
The default method is <code class="literal">btree</code>.
</p></dd><dt><span class="term"><em class="replaceable"><code>column_name</code></em></span></dt><dd><p>
The name of a column of the table.
</p></dd><dt><span class="term"><em class="replaceable"><code>expression</code></em></span></dt><dd><p>
An expression based on one or more columns of the table. The
expression usually must be written with surrounding parentheses,
as shown in the syntax. However, the parentheses can be omitted
if the expression has the form of a function call.
</p></dd><dt><span class="term"><em class="replaceable"><code>collation</code></em></span></dt><dd><p>
The name of the collation to use for the index. By default,
the index uses the collation declared for the column to be
indexed or the result collation of the expression to be
indexed. Indexes with non-default collations can be useful for
queries that involve expressions using non-default collations.
</p></dd><dt><span class="term"><em class="replaceable"><code>opclass</code></em></span></dt><dd><p>
The name of an operator class. See below for details.
</p></dd><dt><span class="term"><em class="replaceable"><code>opclass_parameter</code></em></span></dt><dd><p>
The name of an operator class parameter. See below for details.
</p></dd><dt><span class="term"><code class="literal">ASC</code></span></dt><dd><p>
Specifies ascending sort order (which is the default).
</p></dd><dt><span class="term"><code class="literal">DESC</code></span></dt><dd><p>
Specifies descending sort order.
</p></dd><dt><span class="term"><code class="literal">NULLS FIRST</code></span></dt><dd><p>
Specifies that nulls sort before non-nulls. This is the default
when <code class="literal">DESC</code> is specified.
</p></dd><dt><span class="term"><code class="literal">NULLS LAST</code></span></dt><dd><p>
Specifies that nulls sort after non-nulls. This is the default
when <code class="literal">DESC</code> is not specified.
</p></dd><dt><span class="term"><code class="literal">NULLS DISTINCT</code><br /></span><span class="term"><code class="literal">NULLS NOT DISTINCT</code></span></dt><dd><p>
Specifies whether for a unique index, null values should be considered
distinct (not equal). The default is that they are distinct, so that
a unique index could contain multiple null values in a column.
</p></dd><dt><span class="term"><em class="replaceable"><code>storage_parameter</code></em></span></dt><dd><p>
The name of an index-method-specific storage parameter. See
<a class="xref" href="sql-createindex.html#SQL-CREATEINDEX-STORAGE-PARAMETERS" title="Index Storage Parameters">Index Storage Parameters</a> below
for details.
</p></dd><dt><span class="term"><em class="replaceable"><code>tablespace_name</code></em></span></dt><dd><p>
The tablespace in which to create the index. If not specified,
<a class="xref" href="runtime-config-client.html#GUC-DEFAULT-TABLESPACE">default_tablespace</a> is consulted, or
<a class="xref" href="runtime-config-client.html#GUC-TEMP-TABLESPACES">temp_tablespaces</a> for indexes on temporary
tables.
</p></dd><dt><span class="term"><em class="replaceable"><code>predicate</code></em></span></dt><dd><p>
The constraint expression for a partial index.
</p></dd></dl></div><div class="refsect2" id="SQL-CREATEINDEX-STORAGE-PARAMETERS"><h3>Index Storage Parameters</h3><p>
The optional <code class="literal">WITH</code> clause specifies <em class="firstterm">storage
parameters</em> for the index. Each index method has its own set of allowed
storage parameters. The B-tree, hash, GiST and SP-GiST index methods all
accept this parameter:
</p><div class="variablelist"><dl class="variablelist"><dt id="INDEX-RELOPTION-FILLFACTOR"><span class="term"><code class="literal">fillfactor</code> (<code class="type">integer</code>)
<a id="id-1.9.3.69.6.3.3.1.1.3" class="indexterm"></a>
</span></dt><dd><p>
The fillfactor for an index is a percentage that determines how full
the index method will try to pack index pages. For B-trees, leaf pages
are filled to this percentage during initial index builds, and also
when extending the index at the right (adding new largest key values).
If pages
subsequently become completely full, they will be split, leading to
fragmentation of the on-disk index structure. B-trees use a default
fillfactor of 90, but any integer value from 10 to 100 can be selected.
</p><p>
B-tree indexes on tables where many inserts and/or updates are
anticipated can benefit from lower fillfactor settings at
<code class="command">CREATE INDEX</code> time (following bulk loading into the
table). Values in the range of 50 - 90 can usefully <span class="quote">“<span class="quote">smooth
out</span>”</span> the <span class="emphasis"><em>rate</em></span> of page splits during the
early life of the B-tree index (lowering fillfactor like this may even
lower the absolute number of page splits, though this effect is highly
workload dependent). The B-tree bottom-up index deletion technique
described in <a class="xref" href="btree-implementation.html#BTREE-DELETION" title="67.4.2. Bottom-up Index Deletion">Section 67.4.2</a> is dependent on having
some <span class="quote">“<span class="quote">extra</span>”</span> space on pages to store <span class="quote">“<span class="quote">extra</span>”</span>
tuple versions, and so can be affected by fillfactor (though the effect
is usually not significant).
</p><p>
In other specific cases it might be useful to increase fillfactor to
100 at <code class="command">CREATE INDEX</code> time as a way of maximizing
space utilization. You should only consider this when you are
completely sure that the table is static (i.e. that it will never be
affected by either inserts or updates). A fillfactor setting of 100
otherwise risks <span class="emphasis"><em>harming</em></span> performance: even a few
updates or inserts will cause a sudden flood of page splits.
</p><p>
The other index methods use fillfactor in different but roughly
analogous ways; the default fillfactor varies between methods.
</p></dd></dl></div><p>
B-tree indexes additionally accept this parameter:
</p><div class="variablelist"><dl class="variablelist"><dt id="INDEX-RELOPTION-DEDUPLICATE-ITEMS"><span class="term"><code class="literal">deduplicate_items</code> (<code class="type">boolean</code>)
<a id="id-1.9.3.69.6.3.5.1.1.3" class="indexterm"></a>
</span></dt><dd><p>
Controls usage of the B-tree deduplication technique described
in <a class="xref" href="btree-implementation.html#BTREE-DEDUPLICATION" title="67.4.3. Deduplication">Section 67.4.3</a>. Set to
<code class="literal">ON</code> or <code class="literal">OFF</code> to enable or
disable the optimization. (Alternative spellings of
<code class="literal">ON</code> and <code class="literal">OFF</code> are allowed as
described in <a class="xref" href="config-setting.html" title="20.1. Setting Parameters">Section 20.1</a>.) The default is
<code class="literal">ON</code>.
</p><div class="note"><h3 class="title">Note</h3><p>
Turning <code class="literal">deduplicate_items</code> off via
<code class="command">ALTER INDEX</code> prevents future insertions from
triggering deduplication, but does not in itself make existing
posting list tuples use the standard tuple representation.
</p></div></dd></dl></div><p>
GiST indexes additionally accept this parameter:
</p><div class="variablelist"><dl class="variablelist"><dt id="INDEX-RELOPTION-BUFFERING"><span class="term"><code class="literal">buffering</code> (<code class="type">enum</code>)
<a id="id-1.9.3.69.6.3.7.1.1.3" class="indexterm"></a>
</span></dt><dd><p>
Determines whether the buffered build technique described in
<a class="xref" href="gist-implementation.html#GIST-BUFFERING-BUILD" title="68.4.1. GiST Index Build Methods">Section 68.4.1</a> is used to build the index. With
<code class="literal">OFF</code> buffering is disabled, with <code class="literal">ON</code>
it is enabled, and with <code class="literal">AUTO</code> it is initially disabled,
but is turned on on-the-fly once the index size reaches
<a class="xref" href="runtime-config-query.html#GUC-EFFECTIVE-CACHE-SIZE">effective_cache_size</a>. The default
is <code class="literal">AUTO</code>.
Note that if sorted build is possible, it will be used instead of
buffered build unless <code class="literal">buffering=ON</code> is specified.
</p></dd></dl></div><p>
GIN indexes accept different parameters:
</p><div class="variablelist"><dl class="variablelist"><dt id="INDEX-RELOPTION-FASTUPDATE"><span class="term"><code class="literal">fastupdate</code> (<code class="type">boolean</code>)
<a id="id-1.9.3.69.6.3.9.1.1.3" class="indexterm"></a>
</span></dt><dd><p>
This setting controls usage of the fast update technique described in
<a class="xref" href="gin-implementation.html#GIN-FAST-UPDATE" title="70.4.1. GIN Fast Update Technique">Section 70.4.1</a>. It is a Boolean parameter:
<code class="literal">ON</code> enables fast update, <code class="literal">OFF</code> disables it.
The default is <code class="literal">ON</code>.
</p><div class="note"><h3 class="title">Note</h3><p>
Turning <code class="literal">fastupdate</code> off via <code class="command">ALTER INDEX</code> prevents
future insertions from going into the list of pending index entries,
but does not in itself flush previous entries. You might want to
<code class="command">VACUUM</code> the table or call <code class="function">gin_clean_pending_list</code>
function afterward to ensure the pending list is emptied.
</p></div></dd></dl></div><div class="variablelist"><dl class="variablelist"><dt id="INDEX-RELOPTION-GIN-PENDING-LIST-LIMIT"><span class="term"><code class="literal">gin_pending_list_limit</code> (<code class="type">integer</code>)
<a id="id-1.9.3.69.6.3.10.1.1.3" class="indexterm"></a>
</span></dt><dd><p>
Custom <a class="xref" href="runtime-config-client.html#GUC-GIN-PENDING-LIST-LIMIT">gin_pending_list_limit</a> parameter.
This value is specified in kilobytes.
</p></dd></dl></div><p>
<acronym class="acronym">BRIN</acronym> indexes accept different parameters:
</p><div class="variablelist"><dl class="variablelist"><dt id="INDEX-RELOPTION-PAGES-PER-RANGE"><span class="term"><code class="literal">pages_per_range</code> (<code class="type">integer</code>)
<a id="id-1.9.3.69.6.3.12.1.1.3" class="indexterm"></a>
</span></dt><dd><p>
Defines the number of table blocks that make up one block range for
each entry of a <acronym class="acronym">BRIN</acronym> index (see <a class="xref" href="brin-intro.html" title="71.1. Introduction">Section 71.1</a>
for more details). The default is <code class="literal">128</code>.
</p></dd><dt id="INDEX-RELOPTION-AUTOSUMMARIZE"><span class="term"><code class="literal">autosummarize</code> (<code class="type">boolean</code>)
<a id="id-1.9.3.69.6.3.12.2.1.3" class="indexterm"></a>
</span></dt><dd><p>
Defines whether a summarization run is queued for the previous page
range whenever an insertion is detected on the next one.
See <a class="xref" href="brin-intro.html#BRIN-OPERATION" title="71.1.1. Index Maintenance">Section 71.1.1</a> for more details.
The default is <code class="literal">off</code>.
</p></dd></dl></div></div><div class="refsect2" id="SQL-CREATEINDEX-CONCURRENTLY"><h3>Building Indexes Concurrently</h3><a id="id-1.9.3.69.6.4.2" class="indexterm"></a><p>
Creating an index can interfere with regular operation of a database.
Normally <span class="productname">PostgreSQL</span> locks the table to be indexed against
writes and performs the entire index build with a single scan of the
table. Other transactions can still read the table, but if they try to
insert, update, or delete rows in the table they will block until the
index build is finished. This could have a severe effect if the system is
a live production database. Very large tables can take many hours to be
indexed, and even for smaller tables, an index build can lock out writers
for periods that are unacceptably long for a production system.
</p><p>
<span class="productname">PostgreSQL</span> supports building indexes without locking
out writes. This method is invoked by specifying the
<code class="literal">CONCURRENTLY</code> option of <code class="command">CREATE INDEX</code>.
When this option is used,
<span class="productname">PostgreSQL</span> must perform two scans of the table, and in
addition it must wait for all existing transactions that could potentially
modify or use the index to terminate. Thus
this method requires more total work than a standard index build and takes
significantly longer to complete. However, since it allows normal
operations to continue while the index is built, this method is useful for
adding new indexes in a production environment. Of course, the extra CPU
and I/O load imposed by the index creation might slow other operations.
</p><p>
In a concurrent index build, the index is actually entered as an
<span class="quote">“<span class="quote">invalid</span>”</span> index into
the system catalogs in one transaction, then two table scans occur in
two more transactions. Before each table scan, the index build must
wait for existing transactions that have modified the table to terminate.
After the second scan, the index build must wait for any transactions
that have a snapshot (see <a class="xref" href="mvcc.html" title="Chapter 13. Concurrency Control">Chapter 13</a>) predating the second
scan to terminate, including transactions used by any phase of concurrent
index builds on other tables, if the indexes involved are partial or have
columns that are not simple column references.
Then finally the index can be marked <span class="quote">“<span class="quote">valid</span>”</span> and ready for use,
and the <code class="command">CREATE INDEX</code> command terminates.
Even then, however, the index may not be immediately usable for queries:
in the worst case, it cannot be used as long as transactions exist that
predate the start of the index build.
</p><p>
If a problem arises while scanning the table, such as a deadlock or a
uniqueness violation in a unique index, the <code class="command">CREATE INDEX</code>
command will fail but leave behind an <span class="quote">“<span class="quote">invalid</span>”</span> index. This index
will be ignored for querying purposes because it might be incomplete;
however it will still consume update overhead. The <span class="application">psql</span>
<code class="command">\d</code> command will report such an index as <code class="literal">INVALID</code>:
</p><pre class="programlisting">
postgres=# \d tab
Table "public.tab"
Column | Type | Collation | Nullable | Default
--------+---------+-----------+----------+---------
col | integer | | |
Indexes:
"idx" btree (col) INVALID
</pre><p>
The recommended recovery
method in such cases is to drop the index and try again to perform
<code class="command">CREATE INDEX CONCURRENTLY</code>. (Another possibility is
to rebuild the index with <code class="command">REINDEX INDEX CONCURRENTLY</code>).
</p><p>
Another caveat when building a unique index concurrently is that the
uniqueness constraint is already being enforced against other transactions
when the second table scan begins. This means that constraint violations
could be reported in other queries prior to the index becoming available
for use, or even in cases where the index build eventually fails. Also,
if a failure does occur in the second scan, the <span class="quote">“<span class="quote">invalid</span>”</span> index
continues to enforce its uniqueness constraint afterwards.
</p><p>
Concurrent builds of expression indexes and partial indexes are supported.
Errors occurring in the evaluation of these expressions could cause
behavior similar to that described above for unique constraint violations.
</p><p>
Regular index builds permit other regular index builds on the
same table to occur simultaneously, but only one concurrent index build
can occur on a table at a time. In either case, schema modification of the
table is not allowed while the index is being built. Another difference is
that a regular <code class="command">CREATE INDEX</code> command can be performed
within a transaction block, but <code class="command">CREATE INDEX CONCURRENTLY</code>
cannot.
</p><p>
Concurrent builds for indexes on partitioned tables are currently not
supported. However, you may concurrently build the index on each
partition individually and then finally create the partitioned index
non-concurrently in order to reduce the time where writes to the
partitioned table will be locked out. In this case, building the
partitioned index is a metadata only operation.
</p></div></div><div class="refsect1" id="id-1.9.3.69.7"><h2>Notes</h2><p>
See <a class="xref" href="indexes.html" title="Chapter 11. Indexes">Chapter 11</a> for information about when indexes can
be used, when they are not used, and in which particular situations
they can be useful.
</p><p>
Currently, only the B-tree, GiST, GIN, and BRIN index methods support
multiple-key-column indexes. Whether there can be multiple key
columns is independent of whether <code class="literal">INCLUDE</code> columns
can be added to the index. Indexes can have up to 32 columns,
including <code class="literal">INCLUDE</code> columns.
(This limit can be altered when building
<span class="productname">PostgreSQL</span>.) Only B-tree currently
supports unique indexes.
</p><p>
An <em class="firstterm">operator class</em> with optional parameters
can be specified for each column of an index.
The operator class identifies the operators to be
used by the index for that column. For example, a B-tree index on
four-byte integers would use the <code class="literal">int4_ops</code> class;
this operator class includes comparison functions for four-byte
integers. In practice the default operator class for the column's data
type is usually sufficient. The main point of having operator classes
is that for some data types, there could be more than one meaningful
ordering. For example, we might want to sort a complex-number data
type either by absolute value or by real part. We could do this by
defining two operator classes for the data type and then selecting
the proper class when creating an index. More information about
operator classes is in <a class="xref" href="indexes-opclass.html" title="11.10. Operator Classes and Operator Families">Section 11.10</a> and in <a class="xref" href="xindex.html" title="38.16. Interfacing Extensions to Indexes">Section 38.16</a>.
</p><p>
When <code class="literal">CREATE INDEX</code> is invoked on a partitioned
table, the default behavior is to recurse to all partitions to ensure
they all have matching indexes.
Each partition is first checked to determine whether an equivalent
index already exists, and if so, that index will become attached as a
partition index to the index being created, which will become its
parent index.
If no matching index exists, a new index will be created and
automatically attached; the name of the new index in each partition
will be determined as if no index name had been specified in the
command.
If the <code class="literal">ONLY</code> option is specified, no recursion
is done, and the index is marked invalid.
(<code class="command">ALTER INDEX ... ATTACH PARTITION</code> marks the index
valid, once all partitions acquire matching indexes.) Note, however,
that any partition that is created in the future using
<code class="command">CREATE TABLE ... PARTITION OF</code> will automatically
have a matching index, regardless of whether <code class="literal">ONLY</code> is
specified.
</p><p>
For index methods that support ordered scans (currently, only B-tree),
the optional clauses <code class="literal">ASC</code>, <code class="literal">DESC</code>, <code class="literal">NULLS
FIRST</code>, and/or <code class="literal">NULLS LAST</code> can be specified to modify
the sort ordering of the index. Since an ordered index can be
scanned either forward or backward, it is not normally useful to create a
single-column <code class="literal">DESC</code> index — that sort ordering is already
available with a regular index. The value of these options is that
multicolumn indexes can be created that match the sort ordering requested
by a mixed-ordering query, such as <code class="literal">SELECT ... ORDER BY x ASC, y
DESC</code>. The <code class="literal">NULLS</code> options are useful if you need to support
<span class="quote">“<span class="quote">nulls sort low</span>”</span> behavior, rather than the default <span class="quote">“<span class="quote">nulls
sort high</span>”</span>, in queries that depend on indexes to avoid sorting steps.
</p><p>
The system regularly collects statistics on all of a table's
columns. Newly-created non-expression indexes can immediately
use these statistics to determine an index's usefulness.
For new expression indexes, it is necessary to run <a class="link" href="sql-analyze.html" title="ANALYZE"><code class="command">ANALYZE</code></a> or wait for
the <a class="link" href="routine-vacuuming.html#AUTOVACUUM" title="25.1.6. The Autovacuum Daemon">autovacuum daemon</a> to analyze
the table to generate statistics for these indexes.
</p><p>
For most index methods, the speed of creating an index is
dependent on the setting of <a class="xref" href="runtime-config-resource.html#GUC-MAINTENANCE-WORK-MEM">maintenance_work_mem</a>.
Larger values will reduce the time needed for index creation, so long
as you don't make it larger than the amount of memory really available,
which would drive the machine into swapping.
</p><p>
<span class="productname">PostgreSQL</span> can build indexes while
leveraging multiple CPUs in order to process the table rows faster.
This feature is known as <em class="firstterm">parallel index
build</em>. For index methods that support building indexes
in parallel (currently, only B-tree),
<code class="varname">maintenance_work_mem</code> specifies the maximum
amount of memory that can be used by each index build operation as
a whole, regardless of how many worker processes were started.
Generally, a cost model automatically determines how many worker
processes should be requested, if any.
</p><p>
Parallel index builds may benefit from increasing
<code class="varname">maintenance_work_mem</code> where an equivalent serial
index build will see little or no benefit. Note that
<code class="varname">maintenance_work_mem</code> may influence the number of
worker processes requested, since parallel workers must have at
least a <code class="literal">32MB</code> share of the total
<code class="varname">maintenance_work_mem</code> budget. There must also be
a remaining <code class="literal">32MB</code> share for the leader process.
Increasing <a class="xref" href="runtime-config-resource.html#GUC-MAX-PARALLEL-MAINTENANCE-WORKERS">max_parallel_maintenance_workers</a>
may allow more workers to be used, which will reduce the time
needed for index creation, so long as the index build is not
already I/O bound. Of course, there should also be sufficient
CPU capacity that would otherwise lie idle.
</p><p>
Setting a value for <code class="literal">parallel_workers</code> via <a class="link" href="sql-altertable.html" title="ALTER TABLE"><code class="command">ALTER TABLE</code></a> directly controls how many parallel
worker processes will be requested by a <code class="command">CREATE
INDEX</code> against the table. This bypasses the cost model
completely, and prevents <code class="varname">maintenance_work_mem</code>
from affecting how many parallel workers are requested. Setting
<code class="literal">parallel_workers</code> to 0 via <code class="command">ALTER
TABLE</code> will disable parallel index builds on the table in
all cases.
</p><div class="tip"><h3 class="title">Tip</h3><p>
You might want to reset <code class="literal">parallel_workers</code> after
setting it as part of tuning an index build. This avoids
inadvertent changes to query plans, since
<code class="literal">parallel_workers</code> affects
<span class="emphasis"><em>all</em></span> parallel table scans.
</p></div><p>
While <code class="command">CREATE INDEX</code> with the
<code class="literal">CONCURRENTLY</code> option supports parallel builds
without special restrictions, only the first table scan is actually
performed in parallel.
</p><p>
Use <a class="link" href="sql-dropindex.html" title="DROP INDEX"><code class="command">DROP INDEX</code></a>
to remove an index.
</p><p>
Like any long-running transaction, <code class="command">CREATE INDEX</code> on a
table can affect which tuples can be removed by concurrent
<code class="command">VACUUM</code> on any other table.
</p><p>
Prior releases of <span class="productname">PostgreSQL</span> also had an
R-tree index method. This method has been removed because
it had no significant advantages over the GiST method.
If <code class="literal">USING rtree</code> is specified, <code class="command">CREATE INDEX</code>
will interpret it as <code class="literal">USING gist</code>, to simplify conversion
of old databases to GiST.
</p><p>
Each backend running <code class="command">CREATE INDEX</code> will report its
progress in the <code class="structname">pg_stat_progress_create_index</code>
view. See <a class="xref" href="progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING" title="28.4.2. CREATE INDEX Progress Reporting">Section 28.4.2</a> for details.
</p></div><div class="refsect1" id="id-1.9.3.69.8"><h2>Examples</h2><p>
To create a unique B-tree index on the column <code class="literal">title</code> in
the table <code class="literal">films</code>:
</p><pre class="programlisting">
CREATE UNIQUE INDEX title_idx ON films (title);
</pre><p>
</p><p>
To create a unique B-tree index on the column <code class="literal">title</code>
with included columns <code class="literal">director</code>
and <code class="literal">rating</code> in the table <code class="literal">films</code>:
</p><pre class="programlisting">
CREATE UNIQUE INDEX title_idx ON films (title) INCLUDE (director, rating);
</pre><p>
</p><p>
To create a B-Tree index with deduplication disabled:
</p><pre class="programlisting">
CREATE INDEX title_idx ON films (title) WITH (deduplicate_items = off);
</pre><p>
</p><p>
To create an index on the expression <code class="literal">lower(title)</code>,
allowing efficient case-insensitive searches:
</p><pre class="programlisting">
CREATE INDEX ON films ((lower(title)));
</pre><p>
(In this example we have chosen to omit the index name, so the system
will choose a name, typically <code class="literal">films_lower_idx</code>.)
</p><p>
To create an index with non-default collation:
</p><pre class="programlisting">
CREATE INDEX title_idx_german ON films (title COLLATE "de_DE");
</pre><p>
</p><p>
To create an index with non-default sort ordering of nulls:
</p><pre class="programlisting">
CREATE INDEX title_idx_nulls_low ON films (title NULLS FIRST);
</pre><p>
</p><p>
To create an index with non-default fill factor:
</p><pre class="programlisting">
CREATE UNIQUE INDEX title_idx ON films (title) WITH (fillfactor = 70);
</pre><p>
</p><p>
To create a <acronym class="acronym">GIN</acronym> index with fast updates disabled:
</p><pre class="programlisting">
CREATE INDEX gin_idx ON documents_table USING GIN (locations) WITH (fastupdate = off);
</pre><p>
</p><p>
To create an index on the column <code class="literal">code</code> in the table
<code class="literal">films</code> and have the index reside in the tablespace
<code class="literal">indexspace</code>:
</p><pre class="programlisting">
CREATE INDEX code_idx ON films (code) TABLESPACE indexspace;
</pre><p>
</p><p>
To create a GiST index on a point attribute so that we
can efficiently use box operators on the result of the
conversion function:
</p><pre class="programlisting">
CREATE INDEX pointloc
ON points USING gist (box(location,location));
SELECT * FROM points
WHERE box(location,location) && '(0,0),(1,1)'::box;
</pre><p>
</p><p>
To create an index without locking out writes to the table:
</p><pre class="programlisting">
CREATE INDEX CONCURRENTLY sales_quantity_index ON sales_table (quantity);
</pre></div><div class="refsect1" id="id-1.9.3.69.9"><h2>Compatibility</h2><p>
<code class="command">CREATE INDEX</code> is a
<span class="productname">PostgreSQL</span> language extension. There
are no provisions for indexes in the SQL standard.
</p></div><div class="refsect1" id="id-1.9.3.69.10"><h2>See Also</h2><span class="simplelist"><a class="xref" href="sql-alterindex.html" title="ALTER INDEX"><span class="refentrytitle">ALTER INDEX</span></a>, <a class="xref" href="sql-dropindex.html" title="DROP INDEX"><span class="refentrytitle">DROP INDEX</span></a>, <a class="xref" href="sql-reindex.html" title="REINDEX"><span class="refentrytitle">REINDEX</span></a>, <a class="xref" href="progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING" title="28.4.2. CREATE INDEX Progress Reporting">Section 28.4.2</a></span></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="sql-creategroup.html" title="CREATE GROUP">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="sql-commands.html" title="SQL Commands">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="sql-createlanguage.html" title="CREATE LANGUAGE">Next</a></td></tr><tr><td width="40%" align="left" valign="top">CREATE GROUP </td><td width="20%" align="center"><a accesskey="h" href="index.html" title="PostgreSQL 15.6 Documentation">Home</a></td><td width="40%" align="right" valign="top"> CREATE LANGUAGE</td></tr></table></div></body></html>
|
