1 <?xml version="1.0" encoding="UTF-8" standalone="no"?>
2 <!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>11.8. Partial Indexes</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="indexes-expressional.html" title="11.7. Indexes on Expressions" /><link rel="next" href="indexes-index-only-scans.html" title="11.9. Index-Only Scans and Covering Indexes" /></head><body id="docContent" class="container-fluid col-10"><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">11.8. Partial Indexes</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="indexes-expressional.html" title="11.7. Indexes on Expressions">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="indexes.html" title="Chapter 11. Indexes">Up</a></td><th width="60%" align="center">Chapter 11. Indexes</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 18.0 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="indexes-index-only-scans.html" title="11.9. Index-Only Scans and Covering Indexes">Next</a></td></tr></table><hr /></div><div class="sect1" id="INDEXES-PARTIAL"><div class="titlepage"><div><div><h2 class="title" style="clear: both">11.8. Partial Indexes <a href="#INDEXES-PARTIAL" class="id_link">#</a></h2></div></div></div><a id="id-1.5.10.11.2" class="indexterm"></a><p>
3 A <em class="firstterm">partial index</em> is an index built over a
4 subset of a table; the subset is defined by a conditional
5 expression (called the <em class="firstterm">predicate</em> of the
6 partial index). The index contains entries only for those table
7 rows that satisfy the predicate. Partial indexes are a specialized
8 feature, but there are several situations in which they are useful.
10 One major reason for using a partial index is to avoid indexing common
11 values. Since a query searching for a common value (one that
12 accounts for more than a few percent of all the table rows) will not
13 use the index anyway, there is no point in keeping those rows in the
14 index at all. This reduces the size of the index, which will speed
15 up those queries that do use the index. It will also speed up many table
16 update operations because the index does not need to be
17 updated in all cases. <a class="xref" href="indexes-partial.html#INDEXES-PARTIAL-EX1" title="Example 11.1. Setting up a Partial Index to Exclude Common Values">Example 11.1</a> shows a
18 possible application of this idea.
19 </p><div class="example" id="INDEXES-PARTIAL-EX1"><p class="title"><strong>Example 11.1. Setting up a Partial Index to Exclude Common Values</strong></p><div class="example-contents"><p>
20 Suppose you are storing web server access logs in a database.
21 Most accesses originate from the IP address range of your organization but
22 some are from elsewhere (say, employees on dial-up connections).
23 If your searches by IP are primarily for outside accesses,
24 you probably do not need to index the IP range that corresponds to your
25 organization's subnet.
27 Assume a table like this:
28 </p><pre class="programlisting">
29 CREATE TABLE access_log (
36 To create a partial index that suits our example, use a command
38 </p><pre class="programlisting">
39 CREATE INDEX access_log_client_ip_ix ON access_log (client_ip)
40 WHERE NOT (client_ip > inet '192.168.100.0' AND
41 client_ip < inet '192.168.100.255');
44 A typical query that can use this index would be:
45 </p><pre class="programlisting">
48 WHERE url = '/index.html' AND client_ip = inet '212.78.10.32';
50 Here the query's IP address is covered by the partial index. The
51 following query cannot use the partial index, as it uses an IP address
52 that is excluded from the index:
53 </p><pre class="programlisting">
56 WHERE url = '/index.html' AND client_ip = inet '192.168.100.23';
59 Observe that this kind of partial index requires that the common
60 values be predetermined, so such partial indexes are best used for
61 data distributions that do not change. Such indexes can be recreated
62 occasionally to adjust for new data distributions, but this adds
64 </p></div></div><br class="example-break" /><p>
65 Another possible use for a partial index is to exclude values from the
67 typical query workload is not interested in; this is shown in <a class="xref" href="indexes-partial.html#INDEXES-PARTIAL-EX2" title="Example 11.2. Setting up a Partial Index to Exclude Uninteresting Values">Example 11.2</a>. This results in the same
68 advantages as listed above, but it prevents the
69 <span class="quote">“<span class="quote">uninteresting</span>”</span> values from being accessed via that
70 index, even if an index scan might be profitable in that
71 case. Obviously, setting up partial indexes for this kind of
72 scenario will require a lot of care and experimentation.
73 </p><div class="example" id="INDEXES-PARTIAL-EX2"><p class="title"><strong>Example 11.2. Setting up a Partial Index to Exclude Uninteresting Values</strong></p><div class="example-contents"><p>
74 If you have a table that contains both billed and unbilled orders,
75 where the unbilled orders take up a small fraction of the total
76 table and yet those are the most-accessed rows, you can improve
77 performance by creating an index on just the unbilled rows. The
78 command to create the index would look like this:
79 </p><pre class="programlisting">
80 CREATE INDEX orders_unbilled_index ON orders (order_nr)
81 WHERE billed is not true;
84 A possible query to use this index would be:
85 </p><pre class="programlisting">
86 SELECT * FROM orders WHERE billed is not true AND order_nr < 10000;
88 However, the index can also be used in queries that do not involve
89 <code class="structfield">order_nr</code> at all, e.g.:
90 </p><pre class="programlisting">
91 SELECT * FROM orders WHERE billed is not true AND amount > 5000.00;
93 This is not as efficient as a partial index on the
94 <code class="structfield">amount</code> column would be, since the system has to
95 scan the entire index. Yet, if there are relatively few unbilled
96 orders, using this partial index just to find the unbilled orders
99 Note that this query cannot use this index:
100 </p><pre class="programlisting">
101 SELECT * FROM orders WHERE order_nr = 3501;
103 The order 3501 might be among the billed or unbilled
105 </p></div></div><br class="example-break" /><p>
106 <a class="xref" href="indexes-partial.html#INDEXES-PARTIAL-EX2" title="Example 11.2. Setting up a Partial Index to Exclude Uninteresting Values">Example 11.2</a> also illustrates that the
107 indexed column and the column used in the predicate do not need to
108 match. <span class="productname">PostgreSQL</span> supports partial
109 indexes with arbitrary predicates, so long as only columns of the
110 table being indexed are involved. However, keep in mind that the
111 predicate must match the conditions used in the queries that
112 are supposed to benefit from the index. To be precise, a partial
113 index can be used in a query only if the system can recognize that
114 the <code class="literal">WHERE</code> condition of the query mathematically implies
115 the predicate of the index.
116 <span class="productname">PostgreSQL</span> does not have a sophisticated
117 theorem prover that can recognize mathematically equivalent
118 expressions that are written in different forms. (Not
119 only is such a general theorem prover extremely difficult to
120 create, it would probably be too slow to be of any real use.)
121 The system can recognize simple inequality implications, for example
122 <span class="quote">“<span class="quote">x < 1</span>”</span> implies <span class="quote">“<span class="quote">x < 2</span>”</span>; otherwise
123 the predicate condition must exactly match part of the query's
124 <code class="literal">WHERE</code> condition
125 or the index will not be recognized as usable. Matching takes
126 place at query planning time, not at run time. As a result,
127 parameterized query clauses do not work with a partial index. For
128 example a prepared query with a parameter might specify
129 <span class="quote">“<span class="quote">x < ?</span>”</span> which will never imply
130 <span class="quote">“<span class="quote">x < 2</span>”</span> for all possible values of the parameter.
132 A third possible use for partial indexes does not require the
133 index to be used in queries at all. The idea here is to create
134 a unique index over a subset of a table, as in <a class="xref" href="indexes-partial.html#INDEXES-PARTIAL-EX3" title="Example 11.3. Setting up a Partial Unique Index">Example 11.3</a>. This enforces uniqueness
135 among the rows that satisfy the index predicate, without constraining
137 </p><div class="example" id="INDEXES-PARTIAL-EX3"><p class="title"><strong>Example 11.3. Setting up a Partial Unique Index</strong></p><div class="example-contents"><p>
138 Suppose that we have a table describing test outcomes. We wish
139 to ensure that there is only one <span class="quote">“<span class="quote">successful</span>”</span> entry for
140 a given subject and target combination, but there might be any number of
141 <span class="quote">“<span class="quote">unsuccessful</span>”</span> entries. Here is one way to do it:
142 </p><pre class="programlisting">
150 CREATE UNIQUE INDEX tests_success_constraint ON tests (subject, target)
153 This is a particularly efficient approach when there are few
154 successful tests and many unsuccessful ones. It is also possible to
155 allow only one null in a column by creating a unique partial index
156 with an <code class="literal">IS NULL</code> restriction.
157 </p></div></div><br class="example-break" /><p>
158 Finally, a partial index can also be used to override the system's
159 query plan choices. Also, data sets with peculiar
160 distributions might cause the system to use an index when it really
161 should not. In that case the index can be set up so that it is not
162 available for the offending query. Normally,
163 <span class="productname">PostgreSQL</span> makes reasonable choices about index
164 usage (e.g., it avoids them when retrieving common values, so the
165 earlier example really only saves index size, it is not required to
166 avoid index usage), and grossly incorrect plan choices are cause
169 Keep in mind that setting up a partial index indicates that you
170 know at least as much as the query planner knows, in particular you
171 know when an index might be profitable. Forming this knowledge
172 requires experience and understanding of how indexes in
173 <span class="productname">PostgreSQL</span> work. In most cases, the
174 advantage of a partial index over a regular index will be minimal.
175 There are cases where they are quite counterproductive, as in <a class="xref" href="indexes-partial.html#INDEXES-PARTIAL-EX4" title="Example 11.4. Do Not Use Partial Indexes as a Substitute for Partitioning">Example 11.4</a>.
176 </p><div class="example" id="INDEXES-PARTIAL-EX4"><p class="title"><strong>Example 11.4. Do Not Use Partial Indexes as a Substitute for Partitioning</strong></p><div class="example-contents"><p>
177 You might be tempted to create a large set of non-overlapping partial
180 </p><pre class="programlisting">
181 CREATE INDEX mytable_cat_1 ON mytable (data) WHERE category = 1;
182 CREATE INDEX mytable_cat_2 ON mytable (data) WHERE category = 2;
183 CREATE INDEX mytable_cat_3 ON mytable (data) WHERE category = 3;
185 CREATE INDEX mytable_cat_<em class="replaceable"><code>N</code></em> ON mytable (data) WHERE category = <em class="replaceable"><code>N</code></em>;
188 This is a bad idea! Almost certainly, you'll be better off with a
189 single non-partial index, declared like
191 </p><pre class="programlisting">
192 CREATE INDEX mytable_cat_data ON mytable (category, data);
195 (Put the category column first, for the reasons described in
196 <a class="xref" href="indexes-multicolumn.html" title="11.3. Multicolumn Indexes">Section 11.3</a>.) While a search in this larger
197 index might have to descend through a couple more tree levels than a
198 search in a smaller index, that's almost certainly going to be cheaper
199 than the planner effort needed to select the appropriate one of the
200 partial indexes. The core of the problem is that the system does not
201 understand the relationship among the partial indexes, and will
202 laboriously test each one to see if it's applicable to the current
205 If your table is large enough that a single index really is a bad idea,
206 you should look into using partitioning instead (see
207 <a class="xref" href="ddl-partitioning.html" title="5.12. Table Partitioning">Section 5.12</a>). With that mechanism, the system
208 does understand that the tables and indexes are non-overlapping, so
209 far better performance is possible.
210 </p></div></div><br class="example-break" /><p>
211 More information about partial indexes can be found in <a class="xref" href="biblio.html#STON89B">[ston89b]</a>, <a class="xref" href="biblio.html#OLSON93" title="Partial indexing in POSTGRES: research project">[olson93]</a>, and <a class="xref" href="biblio.html#SESHADRI95">[seshadri95]</a>.
212 </p></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="indexes-expressional.html" title="11.7. Indexes on Expressions">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="indexes.html" title="Chapter 11. Indexes">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="indexes-index-only-scans.html" title="11.9. Index-Only Scans and Covering Indexes">Next</a></td></tr><tr><td width="40%" align="left" valign="top">11.7. Indexes on Expressions </td><td width="20%" align="center"><a accesskey="h" href="index.html" title="PostgreSQL 18.0 Documentation">Home</a></td><td width="40%" align="right" valign="top"> 11.9. Index-Only Scans and Covering Indexes</td></tr></table></div></body></html>