2
.\" Title: CREATE INDEX
3
.\" Author: The PostgreSQL Global Development Group
4
.\" Generator: DocBook XSL Stylesheets v1.75.1 <http://docbook.sf.net/>
6
.\" Manual: PostgreSQL 9.1beta1 Documentation
7
.\" Source: PostgreSQL 9.1beta1
10
.TH "CREATE INDEX" "7" "2011-04-27" "PostgreSQL 9.1beta1" "PostgreSQL 9.1beta1 Documentation"
11
.\" -----------------------------------------------------------------
12
.\" * set default formatting
13
.\" -----------------------------------------------------------------
14
.\" disable hyphenation
16
.\" disable justification (adjust text to left margin only)
18
.\" -----------------------------------------------------------------
19
.\" * MAIN CONTENT STARTS HERE *
20
.\" -----------------------------------------------------------------
22
CREATE_INDEX \- define a new index
27
CREATE [ UNIQUE ] INDEX [ CONCURRENTLY ] [ \fIname\fR ] ON \fItable\fR [ USING \fImethod\fR ]
28
( { \fIcolumn\fR | ( \fIexpression\fR ) } [ COLLATE \fIcollation\fR ] [ \fIopclass\fR ] [ ASC | DESC ] [ NULLS { FIRST | LAST } ] [, \&.\&.\&.] )
29
[ WITH ( \fIstorage_parameter\fR = \fIvalue\fR [, \&.\&.\&. ] ) ]
30
[ TABLESPACE \fItablespace\fR ]
31
[ WHERE \fIpredicate\fR ]
36
constructs an index on the specified column(s) of the specified table\&. Indexes are primarily used to enhance database performance (though inappropriate use can result in slower performance)\&.
38
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\&.
40
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
42
would allow the clause
43
WHERE upper(col) = \(aqJIM\(aq
47
provides the index methods B\-tree, hash, GiST, and GIN\&. Users can also define their own index methods, but that is fairly complicated\&.
53
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
57
to enforce uniqueness over a subset of a table\&. See
58
Section 11.8, \(lqPartial Indexes\(rq, in the documentation
59
for more discussion\&.
61
The expression used in the
63
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
64
WHERE\&. The same restrictions apply to index fields that are expressions\&.
66
All functions and operators used in an index definition must be
67
\(lqimmutable\(rq, 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
69
clause, remember to mark the function immutable when you create it\&.
74
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\&.
79
When this option is used,
81
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\(aqs done\&. There are several caveats to be aware of when using this option \(em see
82
Building Indexes Concurrently\&.
87
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\&. If the name is omitted,
89
chooses a suitable name based on the parent table\(aqs name and the indexed column name(s)\&.
94
The name (possibly schema\-qualified) of the table to be indexed\&.
99
The name of the index method to be used\&. Choices are
103
gin\&. The default method is
109
The name of a column of the table\&.
114
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\&.
119
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\&.
124
The name of an operator class\&. See below for details\&.
129
Specifies ascending sort order (which is the default)\&.
134
Specifies descending sort order\&.
139
Specifies that nulls sort before non\-nulls\&. This is the default when
146
Specifies that nulls sort after non\-nulls\&. This is the default when
151
\fIstorage_parameter\fR
153
The name of an index\-method\-specific storage parameter\&. See
154
Index Storage Parameters
160
The tablespace in which to create the index\&. If not specified,
164
for indexes on temporary tables\&.
169
The constraint expression for a partial index\&.
171
.SS "Index Storage Parameters"
177
for the index\&. Each index method has its own set of allowed storage parameters\&. The B\-tree, hash and GiST index methods all accept a single parameter:
181
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 build, 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 gradual degradation in the index\(aqs efficiency\&. B\-trees use a default fillfactor of 90, but any integer value from 10 to 100 can be selected\&. If the table is static then fillfactor 100 is best to minimize the index\(aqs physical size, but for heavily updated tables a smaller fillfactor is better to minimize the need for page splits\&. The other index methods use fillfactor in different but roughly analogous ways; the default fillfactor varies between methods\&.
184
GIN indexes accept a different parameter:
188
This setting controls usage of the fast update technique described in
189
Section 54.3.1, \(lqGIN Fast Update Technique\(rq, in the documentation\&. It is a Boolean parameter:
193
disables it\&. (Alternative spellings of
197
are allowed as described in
198
Section 18.1, \(lqSetting Parameters\(rq, in the documentation\&.) The default is
205
.nr an-no-space-flag 1
216
prevents future insertions from going into the list of pending index entries, but does not in itself flush previous entries\&. You might want to
218
the table afterward to ensure the pending list is emptied\&.
222
.SS "Building Indexes Concurrently"
223
.\" index: building concurrently
225
Creating an index can interfere with regular operation of a database\&. Normally
227
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\&.
230
supports building indexes without locking out writes\&. This method is invoked by specifying the
233
CREATE INDEX\&. When this option is used,
235
must perform two scans of the table, and in addition it must wait for all existing transactions that could potentially 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\&.
237
In a concurrent index build, the index is actually entered into the system catalogs in one transaction, then the two table scans occur in a second and third transaction\&. If a problem arises while scanning the table, such as a uniqueness violation in a unique index, the
239
command will fail but leave behind an
241
index\&. This index will be ignored for querying purposes because it might be incomplete; however it will still consume update overhead\&. The
243
command will report such an index as
252
Column | Type | Modifiers
253
\-\-\-\-\-\-\-\-+\-\-\-\-\-\-\-\-\-+\-\-\-\-\-\-\-\-\-\-\-
256
"idx" btree (col) INVALID
262
The recommended recovery method in such cases is to drop the index and try again to perform
263
CREATE INDEX CONCURRENTLY\&. (Another possibility is to rebuild the index with
264
REINDEX\&. However, since
266
does not support concurrent builds, this option is unlikely to seem attractive\&.)
268
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
270
index continues to enforce its uniqueness constraint afterwards\&.
272
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\&.
274
Regular index builds permit other regular index builds on the same table to occur in parallel, but only one concurrent index build can occur on a table at a time\&. In both cases, no other types of schema modification on the table are allowed meanwhile\&. Another difference is that a regular
276
command can be performed within a transaction block, but
277
CREATE INDEX CONCURRENTLY
282
Chapter 11, Indexes, in the documentation
283
for information about when indexes can be used, when they are not used, and in which particular situations they can be useful\&.
285
Currently, only the B\-tree, GiST and GIN index methods support multicolumn indexes\&. Up to 32 fields can be specified by default\&. (This limit can be altered when building
286
PostgreSQL\&.) Only B\-tree currently supports unique indexes\&.
290
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
292
class; this operator class includes comparison functions for four\-byte integers\&. In practice the default operator class for the column\(aqs 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 making an index\&. More information about operator classes is in
293
Section 11.9, \(lqOperator Classes and Operator Families\(rq, in the documentation
295
Section 35.14, \(lqInterfacing Extensions To Indexes\(rq, in the documentation\&.
297
For index methods that support ordered scans (currently, only B\-tree), the optional clauses
302
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
304
index \(em 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
305
SELECT \&.\&.\&. ORDER BY x ASC, y DESC\&. The
307
options are useful if you need to support
308
\(lqnulls sort low\(rq
309
behavior, rather than the default
310
\(lqnulls sort high\(rq, in queries that depend on indexes to avoid sorting steps\&.
312
For most index methods, the speed of creating an index is dependent on the setting of
313
maintenance_work_mem\&. Larger values will reduce the time needed for index creation, so long as you don\(aqt make it larger than the amount of memory really available, which would drive the machine into swapping\&. For hash indexes, the value of
315
is also relevant to index creation time:
317
will use one of two different hash index creation methods depending on whether the estimated index size is more or less than
318
\fIeffective_cache_size\fR\&. For best results, make sure that this parameter is also set to something reflective of available memory, and be careful that the sum of
319
\fImaintenance_work_mem\fR
321
\fIeffective_cache_size\fR
322
is less than the machine\(aqs RAM less whatever space is needed by other programs\&.
325
DROP INDEX (\fBDROP_INDEX\fR(7))
326
to remove an index\&.
330
also had an R\-tree index method\&. This method has been removed because it had no significant advantages over the GiST method\&. If
335
USING gist, to simplify conversion of old databases to GiST\&.
338
To create a B\-tree index on the column
347
CREATE UNIQUE INDEX title_idx ON films (title);
353
To create an index on the expression
354
lower(title), allowing efficient case\-insensitive searches:
360
CREATE INDEX ON films ((lower(title)));
366
(In this example we have chosen to omit the index name, so the system will choose a name, typically
369
To create an index with non\-default collation:
375
CREATE INDEX title_idx_german ON films (title COLLATE "de_DE");
381
To create an index with non\-default sort ordering of nulls:
387
CREATE INDEX title_idx_nulls_low ON films (title NULLS FIRST);
393
To create an index with non\-default fill factor:
399
CREATE UNIQUE INDEX title_idx ON films (title) WITH (fillfactor = 70);
407
index with fast updates disabled:
413
CREATE INDEX gin_idx ON documents_table USING gin (locations) WITH (fastupdate = off);
419
To create an index on the column
423
and have the index reside in the tablespace
430
CREATE INDEX code_idx ON films (code) TABLESPACE indexspace;
436
To create a GiST index on a point attribute so that we can efficiently use box operators on the result of the conversion function:
442
CREATE INDEX pointloc
443
ON points USING gist (box(location,location));
445
WHERE box(location,location) && \(aq(0,0),(1,1)\(aq::box;
451
To create an index without locking out writes to the table:
457
CREATE INDEX CONCURRENTLY sales_quantity_index ON sales_table (quantity);
467
language extension\&. There are no provisions for indexes in the SQL standard\&.
469
ALTER INDEX (\fBALTER_INDEX\fR(7)), DROP INDEX (\fBDROP_INDEX\fR(7))