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>pgbench</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="app-pgbasebackup.html" title="pg_basebackup" /><link rel="next" href="app-pgcombinebackup.html" title="pg_combinebackup" /></head><body id="docContent" class="container-fluid col-10"><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center"><span class="application">pgbench</span></th></tr><tr><td width="10%" align="left"><a accesskey="p" href="app-pgbasebackup.html" title="pg_basebackup">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="reference-client.html" title="PostgreSQL Client Applications">Up</a></td><th width="60%" align="center">PostgreSQL Client Applications</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="app-pgcombinebackup.html" title="pg_combinebackup">Next</a></td></tr></table><hr /></div><div class="refentry" id="PGBENCH"><div class="titlepage"></div><a id="id-1.9.4.11.1" class="indexterm"></a><div class="refnamediv"><h2><span class="refentrytitle"><span class="application">pgbench</span></span></h2><p>pgbench — run a benchmark test on <span class="productname">PostgreSQL</span></p></div><div class="refsynopsisdiv"><h2>Synopsis</h2><div class="cmdsynopsis"><p id="id-1.9.4.11.4.1"><code class="command">pgbench</code> <code class="option">-i</code> [<em class="replaceable"><code>option</code></em>...] [<em class="replaceable"><code>dbname</code></em>]</p></div><div class="cmdsynopsis"><p id="id-1.9.4.11.4.2"><code class="command">pgbench</code> [<em class="replaceable"><code>option</code></em>...] [<em class="replaceable"><code>dbname</code></em>]</p></div></div><div class="refsect1" id="id-1.9.4.11.5"><h2>Description</h2><p>
3 <span class="application">pgbench</span> is a simple program for running benchmark
4 tests on <span class="productname">PostgreSQL</span>. It runs the same sequence of SQL
5 commands over and over, possibly in multiple concurrent database sessions,
6 and then calculates the average transaction rate (transactions per second).
7 By default, <span class="application">pgbench</span> tests a scenario that is
8 loosely based on TPC-B, involving five <code class="command">SELECT</code>,
9 <code class="command">UPDATE</code>, and <code class="command">INSERT</code> commands per transaction.
10 However, it is easy to test other cases by writing your own transaction
13 Typical output from <span class="application">pgbench</span> looks like:
15 </p><pre class="screen">
16 transaction type: <builtin: TPC-B (sort of)>
21 maximum number of tries: 1
22 number of transactions per client: 1000
23 number of transactions actually processed: 10000/10000
24 number of failed transactions: 0 (0.000%)
25 latency average = 11.013 ms
26 latency stddev = 7.351 ms
27 initial connection time = 45.758 ms
28 tps = 896.967014 (without initial connection time)
31 The first seven lines report some of the most important parameter
33 The sixth line reports the maximum number of tries for transactions with
34 serialization or deadlock errors (see <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a>
35 for more information).
36 The eighth line reports the number of transactions completed
37 and intended (the latter being just the product of number of clients
38 and number of transactions per client); these will be equal unless the run
39 failed before completion or some SQL command(s) failed. (In
40 <code class="option">-T</code> mode, only the actual number of transactions is printed.)
41 The next line reports the number of failed transactions due to
42 serialization or deadlock errors (see <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a>
43 for more information).
44 The last line reports the number of transactions per second.
46 The default TPC-B-like transaction test requires specific tables to be
47 set up beforehand. <span class="application">pgbench</span> should be invoked with
48 the <code class="option">-i</code> (initialize) option to create and populate these
49 tables. (When you are testing a custom script, you don't need this
50 step, but will instead need to do whatever setup your test needs.)
51 Initialization looks like:
53 </p><pre class="programlisting">
54 pgbench -i [<span class="optional"> <em class="replaceable"><code>other-options</code></em> </span>] <em class="replaceable"><code>dbname</code></em>
57 where <em class="replaceable"><code>dbname</code></em> is the name of the already-created
58 database to test in. (You may also need <code class="option">-h</code>,
59 <code class="option">-p</code>, and/or <code class="option">-U</code> options to specify how to
60 connect to the database server.)
61 </p><div class="caution"><h3 class="title">Caution</h3><p>
62 <code class="literal">pgbench -i</code> creates four tables <code class="structname">pgbench_accounts</code>,
63 <code class="structname">pgbench_branches</code>, <code class="structname">pgbench_history</code>, and
64 <code class="structname">pgbench_tellers</code>,
65 destroying any existing tables of these names.
66 Be very careful to use another database if you have tables having these
69 At the default <span class="quote">“<span class="quote">scale factor</span>”</span> of 1, the tables initially
70 contain this many rows:
71 </p><pre class="screen">
73 ---------------------------------
76 pgbench_accounts 100000
79 You can (and, for most purposes, probably should) increase the number
80 of rows by using the <code class="option">-s</code> (scale factor) option. The
81 <code class="option">-F</code> (fillfactor) option might also be used at this point.
83 Once you have done the necessary setup, you can run your benchmark
84 with a command that doesn't include <code class="option">-i</code>, that is
86 </p><pre class="programlisting">
87 pgbench [<span class="optional"> <em class="replaceable"><code>options</code></em> </span>] <em class="replaceable"><code>dbname</code></em>
90 In nearly all cases, you'll need some options to make a useful test.
91 The most important options are <code class="option">-c</code> (number of clients),
92 <code class="option">-t</code> (number of transactions), <code class="option">-T</code> (time limit),
93 and <code class="option">-f</code> (specify a custom script file).
94 See below for a full list.
95 </p></div><div class="refsect1" id="id-1.9.4.11.6"><h2>Options</h2><p>
96 The following is divided into three subsections. Different options are
97 used during database initialization and while running benchmarks, but some
98 options are useful in both cases.
99 </p><div class="refsect2" id="PGBENCH-INIT-OPTIONS"><h3>Initialization Options</h3><p>
100 <span class="application">pgbench</span> accepts the following command-line
101 initialization arguments:
103 </p><div class="variablelist"><dl class="variablelist"><dt id="PGBENCH-OPTION-DBNAME"><span class="term"><code class="option">[<span class="optional">-d</span>] <em class="replaceable"><code>dbname</code></em></code><br /></span><span class="term"><code class="option">[<span class="optional">--dbname=</span>]<em class="replaceable"><code>dbname</code></em></code></span> <a href="#PGBENCH-OPTION-DBNAME" class="id_link">#</a></dt><dd><p>
104 Specifies the name of the database to test in. If this is
105 not specified, the environment variable
106 <code class="envar">PGDATABASE</code> is used. If that is not set, the
107 user name specified for the connection is used.
108 </p></dd><dt id="PGBENCH-OPTION-INITIALIZE"><span class="term"><code class="option">-i</code><br /></span><span class="term"><code class="option">--initialize</code></span> <a href="#PGBENCH-OPTION-INITIALIZE" class="id_link">#</a></dt><dd><p>
109 Required to invoke initialization mode.
110 </p></dd><dt id="PGBENCH-OPTION-INIT-STEPS"><span class="term"><code class="option">-I <em class="replaceable"><code>init_steps</code></em></code><br /></span><span class="term"><code class="option">--init-steps=<em class="replaceable"><code>init_steps</code></em></code></span> <a href="#PGBENCH-OPTION-INIT-STEPS" class="id_link">#</a></dt><dd><p>
111 Perform just a selected set of the normal initialization steps.
112 <em class="replaceable"><code>init_steps</code></em> specifies the
113 initialization steps to be performed, using one character per step.
114 Each step is invoked in the specified order.
115 The default is <code class="literal">dtgvp</code>.
116 The available steps are:
118 </p><div class="variablelist"><dl class="variablelist"><dt id="PGBENCH-OPTION-INIT-STEPS-D"><span class="term"><code class="literal">d</code> (Drop)</span> <a href="#PGBENCH-OPTION-INIT-STEPS-D" class="id_link">#</a></dt><dd><p>
119 Drop any existing <span class="application">pgbench</span> tables.
120 </p></dd><dt id="PGBENCH-OPTION-INIT-STEPS-T"><span class="term"><code class="literal">t</code> (create Tables)</span> <a href="#PGBENCH-OPTION-INIT-STEPS-T" class="id_link">#</a></dt><dd><p>
121 Create the tables used by the
122 standard <span class="application">pgbench</span> scenario, namely
123 <code class="structname">pgbench_accounts</code>,
124 <code class="structname">pgbench_branches</code>,
125 <code class="structname">pgbench_history</code>, and
126 <code class="structname">pgbench_tellers</code>.
127 </p></dd><dt id="PGBENCH-OPTION-INIT-STEPS-G"><span class="term"><code class="literal">g</code> or <code class="literal">G</code> (Generate data, client-side or server-side)</span> <a href="#PGBENCH-OPTION-INIT-STEPS-G" class="id_link">#</a></dt><dd><p>
128 Generate data and load it into the standard tables,
129 replacing any data already present.
131 With <code class="literal">g</code> (client-side data generation),
132 data is generated in <code class="command">pgbench</code> client and then
133 sent to the server. This uses the client/server bandwidth
134 extensively through a <code class="command">COPY</code>.
135 <code class="command">pgbench</code> uses the <code class="option">FREEZE</code> option
136 to load data into ordinary (non-partition) tables with version 14
137 or later of <span class="productname">PostgreSQL</span> to speed up
138 subsequent <code class="command">VACUUM</code>.
139 Using <code class="literal">g</code> causes logging to
140 print one message every 100,000 rows while generating data for all
143 With <code class="literal">G</code> (server-side data generation),
144 only small queries are sent from the <code class="command">pgbench</code>
145 client and then data is actually generated in the server.
146 No significant bandwidth is required for this variant, but
147 the server will do more work.
148 Using <code class="literal">G</code> causes logging not to print any progress
149 message while generating data.
151 The default initialization behavior uses client-side data
152 generation (equivalent to <code class="literal">g</code>).
153 </p></dd><dt id="PGBENCH-OPTION-INIT-STEPS-V"><span class="term"><code class="literal">v</code> (Vacuum)</span> <a href="#PGBENCH-OPTION-INIT-STEPS-V" class="id_link">#</a></dt><dd><p>
154 Invoke <code class="command">VACUUM</code> on the standard tables.
155 </p></dd><dt id="PGBENCH-OPTION-INIT-STEPS-P"><span class="term"><code class="literal">p</code> (create Primary keys)</span> <a href="#PGBENCH-OPTION-INIT-STEPS-P" class="id_link">#</a></dt><dd><p>
156 Create primary key indexes on the standard tables.
157 </p></dd><dt id="PGBENCH-OPTION-INIT-STEPS-F"><span class="term"><code class="literal">f</code> (create Foreign keys)</span> <a href="#PGBENCH-OPTION-INIT-STEPS-F" class="id_link">#</a></dt><dd><p>
158 Create foreign key constraints between the standard tables.
159 (Note that this step is not performed by default.)
160 </p></dd></dl></div></dd><dt id="PGBENCH-OPTION-FILLFACTOR"><span class="term"><code class="option">-F</code> <em class="replaceable"><code>fillfactor</code></em><br /></span><span class="term"><code class="option">--fillfactor=</code><em class="replaceable"><code>fillfactor</code></em></span> <a href="#PGBENCH-OPTION-FILLFACTOR" class="id_link">#</a></dt><dd><p>
161 Create the <code class="structname">pgbench_accounts</code>,
162 <code class="structname">pgbench_tellers</code> and
163 <code class="structname">pgbench_branches</code> tables with the given fillfactor.
165 </p></dd><dt id="PGBENCH-OPTION-NO-VACUUM-INIT"><span class="term"><code class="option">-n</code><br /></span><span class="term"><code class="option">--no-vacuum</code></span> <a href="#PGBENCH-OPTION-NO-VACUUM-INIT" class="id_link">#</a></dt><dd><p>
166 Perform no vacuuming during initialization.
167 (This option suppresses the <code class="literal">v</code> initialization step,
168 even if it was specified in <code class="option">-I</code>.)
169 </p></dd><dt id="PGBENCH-OPTION-QUIET"><span class="term"><code class="option">-q</code><br /></span><span class="term"><code class="option">--quiet</code></span> <a href="#PGBENCH-OPTION-QUIET" class="id_link">#</a></dt><dd><p>
170 Switch logging to quiet mode, producing only one progress message per 5
171 seconds. The default logging prints one message each 100,000 rows, which
172 often outputs many lines per second (especially on good hardware).
174 This setting has no effect if <code class="literal">G</code> is specified
175 in <code class="option">-I</code>.
176 </p></dd><dt id="PGBENCH-OPTION-SCALE-INIT"><span class="term"><code class="option">-s</code> <em class="replaceable"><code>scale_factor</code></em><br /></span><span class="term"><code class="option">--scale=</code><em class="replaceable"><code>scale_factor</code></em></span> <a href="#PGBENCH-OPTION-SCALE-INIT" class="id_link">#</a></dt><dd><p>
177 Multiply the number of rows generated by the scale factor.
178 For example, <code class="literal">-s 100</code> will create 10,000,000 rows
179 in the <code class="structname">pgbench_accounts</code> table. Default is 1.
180 When the scale is 20,000 or larger, the columns used to
181 hold account identifiers (<code class="structfield">aid</code> columns)
182 will switch to using larger integers (<code class="type">bigint</code>),
183 in order to be big enough to hold the range of account
185 </p></dd><dt id="PGBENCH-OPTION-FOREIGN-KEYS"><span class="term"><code class="option">--foreign-keys</code></span> <a href="#PGBENCH-OPTION-FOREIGN-KEYS" class="id_link">#</a></dt><dd><p>
186 Create foreign key constraints between the standard tables.
187 (This option adds the <code class="literal">f</code> step to the initialization
188 step sequence, if it is not already present.)
189 </p></dd><dt id="PGBENCH-OPTION-INDEX-TABLESPACE"><span class="term"><code class="option">--index-tablespace=<em class="replaceable"><code>index_tablespace</code></em></code></span> <a href="#PGBENCH-OPTION-INDEX-TABLESPACE" class="id_link">#</a></dt><dd><p>
190 Create indexes in the specified tablespace, rather than the default
192 </p></dd><dt id="PGBENCH-OPTION-PARTITION-METHOD"><span class="term"><code class="option">--partition-method=<em class="replaceable"><code>NAME</code></em></code></span> <a href="#PGBENCH-OPTION-PARTITION-METHOD" class="id_link">#</a></dt><dd><p>
193 Create a partitioned <code class="literal">pgbench_accounts</code> table with
194 <em class="replaceable"><code>NAME</code></em> method.
195 Expected values are <code class="literal">range</code> or <code class="literal">hash</code>.
196 This option requires that <code class="option">--partitions</code> is set to non-zero.
197 If unspecified, default is <code class="literal">range</code>.
198 </p></dd><dt id="PGBENCH-OPTION-PARTITIONS"><span class="term"><code class="option">--partitions=<em class="replaceable"><code>NUM</code></em></code></span> <a href="#PGBENCH-OPTION-PARTITIONS" class="id_link">#</a></dt><dd><p>
199 Create a partitioned <code class="literal">pgbench_accounts</code> table with
200 <em class="replaceable"><code>NUM</code></em> partitions of nearly equal size for
201 the scaled number of accounts.
202 Default is <code class="literal">0</code>, meaning no partitioning.
203 </p></dd><dt id="PGBENCH-OPTION-TABLESPACE"><span class="term"><code class="option">--tablespace=<em class="replaceable"><code>tablespace</code></em></code></span> <a href="#PGBENCH-OPTION-TABLESPACE" class="id_link">#</a></dt><dd><p>
204 Create tables in the specified tablespace, rather than the default
206 </p></dd><dt id="PGBENCH-OPTION-UNLOGGED-TABLES"><span class="term"><code class="option">--unlogged-tables</code></span> <a href="#PGBENCH-OPTION-UNLOGGED-TABLES" class="id_link">#</a></dt><dd><p>
207 Create all tables as unlogged tables, rather than permanent tables.
208 </p></dd></dl></div><p>
209 </p></div><div class="refsect2" id="PGBENCH-RUN-OPTIONS"><h3>Benchmarking Options</h3><p>
210 <span class="application">pgbench</span> accepts the following command-line
211 benchmarking arguments:
213 </p><div class="variablelist"><dl class="variablelist"><dt id="PGBENCH-OPTION-BUILTIN"><span class="term"><code class="option">-b</code> <em class="replaceable"><code>scriptname[@weight]</code></em><br /></span><span class="term"><code class="option">--builtin</code>=<em class="replaceable"><code>scriptname[@weight]</code></em></span> <a href="#PGBENCH-OPTION-BUILTIN" class="id_link">#</a></dt><dd><p>
214 Add the specified built-in script to the list of scripts to be executed.
215 Available built-in scripts are: <code class="literal">tpcb-like</code>,
216 <code class="literal">simple-update</code> and <code class="literal">select-only</code>.
217 Unambiguous prefixes of built-in names are accepted.
218 With the special name <code class="literal">list</code>, show the list of built-in scripts
219 and exit immediately.
221 Optionally, write an integer weight after <code class="literal">@</code> to
222 adjust the probability of selecting this script versus other ones.
223 The default weight is 1.
224 See below for details.
225 </p></dd><dt id="PGBENCH-OPTION-CLIENT"><span class="term"><code class="option">-c</code> <em class="replaceable"><code>clients</code></em><br /></span><span class="term"><code class="option">--client=</code><em class="replaceable"><code>clients</code></em></span> <a href="#PGBENCH-OPTION-CLIENT" class="id_link">#</a></dt><dd><p>
226 Number of clients simulated, that is, number of concurrent database
227 sessions. Default is 1.
228 </p></dd><dt id="PGBENCH-OPTION-CONNECT"><span class="term"><code class="option">-C</code><br /></span><span class="term"><code class="option">--connect</code></span> <a href="#PGBENCH-OPTION-CONNECT" class="id_link">#</a></dt><dd><p>
229 Establish a new connection for each transaction, rather than
230 doing it just once per client session.
231 This is useful to measure the connection overhead.
232 </p></dd><dt id="PGBENCH-OPTION-DEFINE"><span class="term"><code class="option">-D</code> <em class="replaceable"><code>varname</code></em><code class="literal">=</code><em class="replaceable"><code>value</code></em><br /></span><span class="term"><code class="option">--define=</code><em class="replaceable"><code>varname</code></em><code class="literal">=</code><em class="replaceable"><code>value</code></em></span> <a href="#PGBENCH-OPTION-DEFINE" class="id_link">#</a></dt><dd><p>
233 Define a variable for use by a custom script (see below).
234 Multiple <code class="option">-D</code> options are allowed.
235 </p></dd><dt id="PGBENCH-OPTION-FILE"><span class="term"><code class="option">-f</code> <em class="replaceable"><code>filename[@weight]</code></em><br /></span><span class="term"><code class="option">--file=</code><em class="replaceable"><code>filename[@weight]</code></em></span> <a href="#PGBENCH-OPTION-FILE" class="id_link">#</a></dt><dd><p>
236 Add a transaction script read from <em class="replaceable"><code>filename</code></em>
237 to the list of scripts to be executed.
239 Optionally, write an integer weight after <code class="literal">@</code> to
240 adjust the probability of selecting this script versus other ones.
241 The default weight is 1.
242 (To use a script file name that includes an <code class="literal">@</code>
243 character, append a weight so that there is no ambiguity, for
244 example <code class="literal">filen@me@1</code>.)
245 See below for details.
246 </p></dd><dt id="PGBENCH-OPTION-JOBS"><span class="term"><code class="option">-j</code> <em class="replaceable"><code>threads</code></em><br /></span><span class="term"><code class="option">--jobs=</code><em class="replaceable"><code>threads</code></em></span> <a href="#PGBENCH-OPTION-JOBS" class="id_link">#</a></dt><dd><p>
247 Number of worker threads within <span class="application">pgbench</span>.
248 Using more than one thread can be helpful on multi-CPU machines.
249 Clients are distributed as evenly as possible among available threads.
251 </p></dd><dt id="PGBENCH-OPTION-LOG"><span class="term"><code class="option">-l</code><br /></span><span class="term"><code class="option">--log</code></span> <a href="#PGBENCH-OPTION-LOG" class="id_link">#</a></dt><dd><p>
252 Write information about each transaction to a log file.
253 See below for details.
254 </p></dd><dt id="PGBENCH-OPTION-LATENCY-LIMIT"><span class="term"><code class="option">-L</code> <em class="replaceable"><code>limit</code></em><br /></span><span class="term"><code class="option">--latency-limit=</code><em class="replaceable"><code>limit</code></em></span> <a href="#PGBENCH-OPTION-LATENCY-LIMIT" class="id_link">#</a></dt><dd><p>
255 Transactions that last more than <em class="replaceable"><code>limit</code></em> milliseconds
256 are counted and reported separately, as <em class="firstterm">late</em>.
258 When throttling is used (<code class="option">--rate=...</code>), transactions that
259 lag behind schedule by more than <em class="replaceable"><code>limit</code></em> ms, and thus
260 have no hope of meeting the latency limit, are not sent to the server
261 at all. They are counted and reported separately as
262 <em class="firstterm">skipped</em>.
264 When the <code class="option">--max-tries</code> option is used, a transaction
265 which fails due to a serialization anomaly or from a deadlock will not
266 be retried if the total time of all its tries is greater than
267 <em class="replaceable"><code>limit</code></em> ms. To limit only the time of tries
268 and not their number, use <code class="literal">--max-tries=0</code>. By
269 default, the option <code class="option">--max-tries</code> is set to 1 and
270 transactions with serialization/deadlock errors are not retried. See
271 <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a> for more information about
272 retrying such transactions.
273 </p></dd><dt id="PGBENCH-OPTION-PROTOCOL"><span class="term"><code class="option">-M</code> <em class="replaceable"><code>querymode</code></em><br /></span><span class="term"><code class="option">--protocol=</code><em class="replaceable"><code>querymode</code></em></span> <a href="#PGBENCH-OPTION-PROTOCOL" class="id_link">#</a></dt><dd><p>
274 Protocol to use for submitting queries to the server:
275 </p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p><code class="literal">simple</code>: use simple query protocol.</p></li><li class="listitem"><p><code class="literal">extended</code>: use extended query protocol.</p></li><li class="listitem"><p><code class="literal">prepared</code>: use extended query protocol with prepared statements.</p></li></ul></div><p>
277 In the <code class="literal">prepared</code> mode, <span class="application">pgbench</span>
278 reuses the parse analysis result starting from the second query
279 iteration, so <span class="application">pgbench</span> runs faster
282 The default is simple query protocol. (See <a class="xref" href="protocol.html" title="Chapter 54. Frontend/Backend Protocol">Chapter 54</a>
283 for more information.)
284 </p></dd><dt id="PGBENCH-OPTION-NO-VACUUM-RUN"><span class="term"><code class="option">-n</code><br /></span><span class="term"><code class="option">--no-vacuum</code></span> <a href="#PGBENCH-OPTION-NO-VACUUM-RUN" class="id_link">#</a></dt><dd><p>
285 Perform no vacuuming before running the test.
286 This option is <span class="emphasis"><em>necessary</em></span>
287 if you are running a custom test scenario that does not include
288 the standard tables <code class="structname">pgbench_accounts</code>,
289 <code class="structname">pgbench_branches</code>, <code class="structname">pgbench_history</code>, and
290 <code class="structname">pgbench_tellers</code>.
291 </p></dd><dt id="PGBENCH-OPTION-SKIP-SOME-UPDATES"><span class="term"><code class="option">-N</code><br /></span><span class="term"><code class="option">--skip-some-updates</code></span> <a href="#PGBENCH-OPTION-SKIP-SOME-UPDATES" class="id_link">#</a></dt><dd><p>
292 Run built-in simple-update script.
293 Shorthand for <code class="option">-b simple-update</code>.
294 </p></dd><dt id="PGBENCH-OPTION-PROGRESS"><span class="term"><code class="option">-P</code> <em class="replaceable"><code>sec</code></em><br /></span><span class="term"><code class="option">--progress=</code><em class="replaceable"><code>sec</code></em></span> <a href="#PGBENCH-OPTION-PROGRESS" class="id_link">#</a></dt><dd><p>
295 Show progress report every <em class="replaceable"><code>sec</code></em> seconds. The report
296 includes the time since the beginning of the run, the TPS since the
297 last report, and the transaction latency average, standard deviation,
298 and the number of failed transactions since the last report. Under
299 throttling (<code class="option">-R</code>), the latency is computed with respect
300 to the transaction scheduled start time, not the actual transaction
301 beginning time, thus it also includes the average schedule lag time.
302 When <code class="option">--max-tries</code> is used to enable transaction retries
303 after serialization/deadlock errors, the report includes the number of
304 retried transactions and the sum of all retries.
305 </p></dd><dt id="PGBENCH-OPTION-REPORT-LATENCIES"><span class="term"><code class="option">-r</code><br /></span><span class="term"><code class="option">--report-per-command</code></span> <a href="#PGBENCH-OPTION-REPORT-LATENCIES" class="id_link">#</a></dt><dd><p>
306 Report the following statistics for each command after the benchmark
307 finishes: the average per-statement latency (execution time from the
308 perspective of the client), the number of failures, and the number of
309 retries after serialization or deadlock errors in this command. The
310 report displays retry statistics only if the
311 <code class="option">--max-tries</code> option is not equal to 1.
312 </p></dd><dt id="PGBENCH-OPTION-RATE"><span class="term"><code class="option">-R</code> <em class="replaceable"><code>rate</code></em><br /></span><span class="term"><code class="option">--rate=</code><em class="replaceable"><code>rate</code></em></span> <a href="#PGBENCH-OPTION-RATE" class="id_link">#</a></dt><dd><p>
313 Execute transactions targeting the specified rate instead of running
314 as fast as possible (the default). The rate is given in transactions
315 per second. If the targeted rate is above the maximum possible rate,
316 the rate limit won't impact the results.
318 The rate is targeted by starting transactions along a
319 Poisson-distributed schedule time line. The expected start time
320 schedule moves forward based on when the client first started, not
321 when the previous transaction ended. That approach means that when
322 transactions go past their original scheduled end time, it is
323 possible for later ones to catch up again.
325 When throttling is active, the transaction latency reported at the
326 end of the run is calculated from the scheduled start times, so it
327 includes the time each transaction had to wait for the previous
328 transaction to finish. The wait time is called the schedule lag time,
329 and its average and maximum are also reported separately. The
330 transaction latency with respect to the actual transaction start time,
331 i.e., the time spent executing the transaction in the database, can be
332 computed by subtracting the schedule lag time from the reported
335 If <code class="option">--latency-limit</code> is used together with <code class="option">--rate</code>,
336 a transaction can lag behind so much that it is already over the
337 latency limit when the previous transaction ends, because the latency
338 is calculated from the scheduled start time. Such transactions are
339 not sent to the server, but are skipped altogether and counted
342 A high schedule lag time is an indication that the system cannot
343 process transactions at the specified rate, with the chosen number of
344 clients and threads. When the average transaction execution time is
345 longer than the scheduled interval between each transaction, each
346 successive transaction will fall further behind, and the schedule lag
347 time will keep increasing the longer the test run is. When that
348 happens, you will have to reduce the specified transaction rate.
349 </p></dd><dt id="PGBENCH-OPTION-SCALE-RUN"><span class="term"><code class="option">-s</code> <em class="replaceable"><code>scale_factor</code></em><br /></span><span class="term"><code class="option">--scale=</code><em class="replaceable"><code>scale_factor</code></em></span> <a href="#PGBENCH-OPTION-SCALE-RUN" class="id_link">#</a></dt><dd><p>
350 Report the specified scale factor in <span class="application">pgbench</span>'s
351 output. With the built-in tests, this is not necessary; the
352 correct scale factor will be detected by counting the number of
353 rows in the <code class="structname">pgbench_branches</code> table.
354 However, when testing only custom benchmarks (<code class="option">-f</code> option),
355 the scale factor will be reported as 1 unless this option is used.
356 </p></dd><dt id="PGBENCH-OPTION-SELECT-ONLY"><span class="term"><code class="option">-S</code><br /></span><span class="term"><code class="option">--select-only</code></span> <a href="#PGBENCH-OPTION-SELECT-ONLY" class="id_link">#</a></dt><dd><p>
357 Run built-in select-only script.
358 Shorthand for <code class="option">-b select-only</code>.
359 </p></dd><dt id="PGBENCH-OPTION-TRANSACTIONS"><span class="term"><code class="option">-t</code> <em class="replaceable"><code>transactions</code></em><br /></span><span class="term"><code class="option">--transactions=</code><em class="replaceable"><code>transactions</code></em></span> <a href="#PGBENCH-OPTION-TRANSACTIONS" class="id_link">#</a></dt><dd><p>
360 Number of transactions each client runs. Default is 10.
361 </p></dd><dt id="PGBENCH-OPTION-TIME"><span class="term"><code class="option">-T</code> <em class="replaceable"><code>seconds</code></em><br /></span><span class="term"><code class="option">--time=</code><em class="replaceable"><code>seconds</code></em></span> <a href="#PGBENCH-OPTION-TIME" class="id_link">#</a></dt><dd><p>
362 Run the test for this many seconds, rather than a fixed number of
363 transactions per client. <code class="option">-t</code> and
364 <code class="option">-T</code> are mutually exclusive.
365 </p></dd><dt id="PGBENCH-OPTION-VACUUM-ALL"><span class="term"><code class="option">-v</code><br /></span><span class="term"><code class="option">--vacuum-all</code></span> <a href="#PGBENCH-OPTION-VACUUM-ALL" class="id_link">#</a></dt><dd><p>
366 Vacuum all four standard tables before running the test.
367 With neither <code class="option">-n</code> nor <code class="option">-v</code>, <span class="application">pgbench</span> will vacuum the
368 <code class="structname">pgbench_tellers</code> and <code class="structname">pgbench_branches</code>
369 tables, and will truncate <code class="structname">pgbench_history</code>.
370 </p></dd><dt id="PGBENCH-OPTION-AGGREGATE-INTERVAL"><span class="term"><code class="option">--aggregate-interval=<em class="replaceable"><code>seconds</code></em></code></span> <a href="#PGBENCH-OPTION-AGGREGATE-INTERVAL" class="id_link">#</a></dt><dd><p>
371 Length of aggregation interval (in seconds). May be used only
372 with <code class="option">-l</code> option. With this option, the log contains
373 per-interval summary data, as described below.
374 </p></dd><dt id="PGBENCH-OPTION-EXIT-ON-ABORT"><span class="term"><code class="option">--exit-on-abort</code></span> <a href="#PGBENCH-OPTION-EXIT-ON-ABORT" class="id_link">#</a></dt><dd><p>
375 Exit immediately when any client is aborted due to some error. Without
376 this option, even when a client is aborted, other clients could
377 continue their run as specified by <code class="option">-t</code>
378 or <code class="option">-T</code> option, and <span class="application">pgbench</span>
379 will print an incomplete results in this case.
381 Note that serialization failures or deadlock failures do not abort the
382 client, so they are not affected by this option.
383 See <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a> for more information.
384 </p></dd><dt id="PGBENCH-OPTION-FAILURES-DETAILED"><span class="term"><code class="option">--failures-detailed</code></span> <a href="#PGBENCH-OPTION-FAILURES-DETAILED" class="id_link">#</a></dt><dd><p>
385 Report failures in per-transaction and aggregation logs, as well as in
386 the main and per-script reports, grouped by the following types:
387 </p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>serialization failures;</p></li><li class="listitem"><p>deadlock failures;</p></li></ul></div><p>
388 See <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a> for more information.
389 </p></dd><dt id="PGBENCH-OPTION-LOG-PREFIX"><span class="term"><code class="option">--log-prefix=<em class="replaceable"><code>prefix</code></em></code></span> <a href="#PGBENCH-OPTION-LOG-PREFIX" class="id_link">#</a></dt><dd><p>
390 Set the filename prefix for the log files created by
391 <code class="option">--log</code>. The default is <code class="literal">pgbench_log</code>.
392 </p></dd><dt id="PGBENCH-OPTION-MAX-TRIES"><span class="term"><code class="option">--max-tries=<em class="replaceable"><code>number_of_tries</code></em></code></span> <a href="#PGBENCH-OPTION-MAX-TRIES" class="id_link">#</a></dt><dd><p>
393 Enable retries for transactions with serialization/deadlock errors and
394 set the maximum number of these tries. This option can be combined with
395 the <code class="option">--latency-limit</code> option which limits the total time
396 of all transaction tries; moreover, you cannot use an unlimited number
397 of tries (<code class="literal">--max-tries=0</code>) without
398 <code class="option">--latency-limit</code> or <code class="option">--time</code>.
399 The default value is 1 and transactions with serialization/deadlock
400 errors are not retried. See <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a>
401 for more information about retrying such transactions.
402 </p></dd><dt id="PGBENCH-OPTION-PROGRESS-TIMESTAMP"><span class="term"><code class="option">--progress-timestamp</code></span> <a href="#PGBENCH-OPTION-PROGRESS-TIMESTAMP" class="id_link">#</a></dt><dd><p>
403 When showing progress (option <code class="option">-P</code>), use a timestamp
404 (Unix epoch) instead of the number of seconds since the
405 beginning of the run. The unit is in seconds, with millisecond
406 precision after the dot.
407 This helps compare logs generated by various tools.
408 </p></dd><dt id="PGBENCH-OPTION-RANDOM-SEED"><span class="term"><code class="option">--random-seed=</code><em class="replaceable"><code>seed</code></em></span> <a href="#PGBENCH-OPTION-RANDOM-SEED" class="id_link">#</a></dt><dd><p>
409 Set random generator seed. Seeds the system random number generator,
410 which then produces a sequence of initial generator states, one for
412 Values for <em class="replaceable"><code>seed</code></em> may be:
413 <code class="literal">time</code> (the default, the seed is based on the current time),
414 <code class="literal">rand</code> (use a strong random source, failing if none
415 is available), or an unsigned decimal integer value.
416 The random generator is invoked explicitly from a pgbench script
417 (<code class="literal">random...</code> functions) or implicitly (for instance option
418 <code class="option">--rate</code> uses it to schedule transactions).
419 When explicitly set, the value used for seeding is shown on the terminal.
420 Any value allowed for <em class="replaceable"><code>seed</code></em> may also be
421 provided through the environment variable
422 <code class="literal">PGBENCH_RANDOM_SEED</code>.
423 To ensure that the provided seed impacts all possible uses, put this option
424 first or use the environment variable.
426 Setting the seed explicitly allows to reproduce a <code class="command">pgbench</code>
427 run exactly, as far as random numbers are concerned.
428 As the random state is managed per thread, this means the exact same
429 <code class="command">pgbench</code> run for an identical invocation if there is one
430 client per thread and there are no external or data dependencies.
431 From a statistical viewpoint reproducing runs exactly is a bad idea because
432 it can hide the performance variability or improve performance unduly,
433 e.g., by hitting the same pages as a previous run.
434 However, it may also be of great help for debugging, for instance
435 re-running a tricky case which leads to an error.
437 </p></dd><dt id="PGBENCH-OPTION-SAMPLING-RATE"><span class="term"><code class="option">--sampling-rate=<em class="replaceable"><code>rate</code></em></code></span> <a href="#PGBENCH-OPTION-SAMPLING-RATE" class="id_link">#</a></dt><dd><p>
438 Sampling rate, used when writing data into the log, to reduce the
439 amount of log generated. If this option is given, only the specified
440 fraction of transactions are logged. 1.0 means all transactions will
441 be logged, 0.05 means only 5% of the transactions will be logged.
443 Remember to take the sampling rate into account when processing the
444 log file. For example, when computing TPS values, you need to multiply
445 the numbers accordingly (e.g., with 0.01 sample rate, you'll only get
446 1/100 of the actual TPS).
447 </p></dd><dt id="PGBENCH-OPTION-SHOW-SCRIPT"><span class="term"><code class="option">--show-script=</code><em class="replaceable"><code>scriptname</code></em></span> <a href="#PGBENCH-OPTION-SHOW-SCRIPT" class="id_link">#</a></dt><dd><p>
448 Show the actual code of builtin script <em class="replaceable"><code>scriptname</code></em>
449 on stderr, and exit immediately.
450 </p></dd><dt id="PGBENCH-OPTION-VERBOSE-ERRORS"><span class="term"><code class="option">--verbose-errors</code></span> <a href="#PGBENCH-OPTION-VERBOSE-ERRORS" class="id_link">#</a></dt><dd><p>
451 Print messages about all errors and failures (errors without retrying)
452 including which limit for retries was exceeded and how far it was
453 exceeded for the serialization/deadlock failures. (Note that in this
454 case the output can be significantly increased.)
455 See <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a> for more information.
456 </p></dd></dl></div><p>
457 </p></div><div class="refsect2" id="PGBENCH-COMMON-OPTIONS"><h3>Common Options</h3><p>
458 <span class="application">pgbench</span> also accepts the following common command-line
459 arguments for connection parameters and other common settings:
461 </p><div class="variablelist"><dl class="variablelist"><dt id="PGBENCH-OPTION-DEBUG"><span class="term"><code class="option">--debug</code></span> <a href="#PGBENCH-OPTION-DEBUG" class="id_link">#</a></dt><dd><p>
462 Print debugging output.
463 </p></dd><dt id="PGBENCH-OPTION-HOST"><span class="term"><code class="option">-h</code> <em class="replaceable"><code>hostname</code></em><br /></span><span class="term"><code class="option">--host=</code><em class="replaceable"><code>hostname</code></em></span> <a href="#PGBENCH-OPTION-HOST" class="id_link">#</a></dt><dd><p>
464 The database server's host name
465 </p></dd><dt id="PGBENCH-OPTION-PORT"><span class="term"><code class="option">-p</code> <em class="replaceable"><code>port</code></em><br /></span><span class="term"><code class="option">--port=</code><em class="replaceable"><code>port</code></em></span> <a href="#PGBENCH-OPTION-PORT" class="id_link">#</a></dt><dd><p>
466 The database server's port number
467 </p></dd><dt id="PGBENCH-OPTION-USERNAME"><span class="term"><code class="option">-U</code> <em class="replaceable"><code>login</code></em><br /></span><span class="term"><code class="option">--username=</code><em class="replaceable"><code>login</code></em></span> <a href="#PGBENCH-OPTION-USERNAME" class="id_link">#</a></dt><dd><p>
468 The user name to connect as
469 </p></dd><dt id="PGBENCH-OPTION-VERSION"><span class="term"><code class="option">-V</code><br /></span><span class="term"><code class="option">--version</code></span> <a href="#PGBENCH-OPTION-VERSION" class="id_link">#</a></dt><dd><p>
470 Print the <span class="application">pgbench</span> version and exit.
471 </p></dd><dt id="PGBENCH-OPTION-HELP"><span class="term"><code class="option">-?</code><br /></span><span class="term"><code class="option">--help</code></span> <a href="#PGBENCH-OPTION-HELP" class="id_link">#</a></dt><dd><p>
472 Show help about <span class="application">pgbench</span> command line
474 </p></dd></dl></div><p>
475 </p></div></div><div class="refsect1" id="id-1.9.4.11.7"><h2>Exit Status</h2><p>
476 A successful run will exit with status 0. Exit status 1 indicates static
477 problems such as invalid command-line options or internal errors which
478 are supposed to never occur. Early errors that occur when starting
479 benchmark such as initial connection failures also exit with status 1.
480 Errors during the run such as database errors or problems in the script
481 will result in exit status 2. In the latter case,
482 <span class="application">pgbench</span> will print partial results if
483 <code class="option">--exit-on-abort</code> option is not specified.
484 </p></div><div class="refsect1" id="id-1.9.4.11.8"><h2>Environment</h2><div class="variablelist"><dl class="variablelist"><dt id="PGBENCH-ENVIRONMENT-PGDATABASE"><span class="term"><code class="envar">PGDATABASE</code><br /></span><span class="term"><code class="envar">PGHOST</code><br /></span><span class="term"><code class="envar">PGPORT</code><br /></span><span class="term"><code class="envar">PGUSER</code></span> <a href="#PGBENCH-ENVIRONMENT-PGDATABASE" class="id_link">#</a></dt><dd><p>
485 Default connection parameters.
486 </p></dd></dl></div><p>
487 This utility, like most other <span class="productname">PostgreSQL</span> utilities,
488 uses the environment variables supported by <span class="application">libpq</span>
489 (see <a class="xref" href="libpq-envars.html" title="32.15. Environment Variables">Section 32.15</a>).
491 The environment variable <code class="envar">PG_COLOR</code> specifies whether to use
492 color in diagnostic messages. Possible values are
493 <code class="literal">always</code>, <code class="literal">auto</code> and
494 <code class="literal">never</code>.
495 </p></div><div class="refsect1" id="id-1.9.4.11.9"><h2>Notes</h2><div class="refsect2" id="TRANSACTIONS-AND-SCRIPTS"><h3>What Is the <span class="quote">“<span class="quote">Transaction</span>”</span> Actually Performed in <span class="application">pgbench</span>?</h3><p>
496 <span class="application">pgbench</span> executes test scripts chosen randomly
497 from a specified list.
498 The scripts may include built-in scripts specified with <code class="option">-b</code>
499 and user-provided scripts specified with <code class="option">-f</code>.
500 Each script may be given a relative weight specified after an
501 <code class="literal">@</code> so as to change its selection probability.
502 The default weight is <code class="literal">1</code>.
503 Scripts with a weight of <code class="literal">0</code> are ignored.
505 The default built-in transaction script (also invoked with <code class="option">-b tpcb-like</code>)
506 issues seven commands per transaction over randomly chosen <code class="literal">aid</code>,
507 <code class="literal">tid</code>, <code class="literal">bid</code> and <code class="literal">delta</code>.
508 The scenario is inspired by the TPC-B benchmark, but is not actually TPC-B,
510 </p><div class="orderedlist"><ol class="orderedlist" type="1"><li class="listitem"><p><code class="literal">BEGIN;</code></p></li><li class="listitem"><p><code class="literal">UPDATE pgbench_accounts SET abalance = abalance + :delta WHERE aid = :aid;</code></p></li><li class="listitem"><p><code class="literal">SELECT abalance FROM pgbench_accounts WHERE aid = :aid;</code></p></li><li class="listitem"><p><code class="literal">UPDATE pgbench_tellers SET tbalance = tbalance + :delta WHERE tid = :tid;</code></p></li><li class="listitem"><p><code class="literal">UPDATE pgbench_branches SET bbalance = bbalance + :delta WHERE bid = :bid;</code></p></li><li class="listitem"><p><code class="literal">INSERT INTO pgbench_history (tid, bid, aid, delta, mtime) VALUES (:tid, :bid, :aid, :delta, CURRENT_TIMESTAMP);</code></p></li><li class="listitem"><p><code class="literal">END;</code></p></li></ol></div><p>
511 If you select the <code class="literal">simple-update</code> built-in (also <code class="option">-N</code>),
512 steps 4 and 5 aren't included in the transaction.
513 This will avoid update contention on these tables, but
514 it makes the test case even less like TPC-B.
516 If you select the <code class="literal">select-only</code> built-in (also <code class="option">-S</code>),
517 only the <code class="command">SELECT</code> is issued.
518 </p></div><div class="refsect2" id="id-1.9.4.11.9.3"><h3>Custom Scripts</h3><p>
519 <span class="application">pgbench</span> has support for running custom
520 benchmark scenarios by replacing the default transaction script
521 (described above) with a transaction script read from a file
522 (<code class="option">-f</code> option). In this case a <span class="quote">“<span class="quote">transaction</span>”</span>
523 counts as one execution of a script file.
525 A script file contains one or more SQL commands terminated by
526 semicolons. Empty lines and lines beginning with
527 <code class="literal">--</code> are ignored. Script files can also contain
528 <span class="quote">“<span class="quote">meta commands</span>”</span>, which are interpreted by <span class="application">pgbench</span>
529 itself, as described below.
530 </p><div class="note"><h3 class="title">Note</h3><p>
531 Before <span class="productname">PostgreSQL</span> 9.6, SQL commands in script files
532 were terminated by newlines, and so they could not be continued across
533 lines. Now a semicolon is <span class="emphasis"><em>required</em></span> to separate consecutive
534 SQL commands (though an SQL command does not need one if it is followed
535 by a meta command). If you need to create a script file that works with
536 both old and new versions of <span class="application">pgbench</span>, be sure to write
537 each SQL command on a single line ending with a semicolon.
539 It is assumed that <span class="application">pgbench</span> scripts do not contain
540 incomplete blocks of SQL
541 transactions. If at runtime the client reaches the end of the script without
542 completing the last transaction block, it will be aborted.
544 There is a simple variable-substitution facility for script files.
545 Variable names must consist of letters (including non-Latin letters),
546 digits, and underscores, with the first character not being a digit.
547 Variables can be set by the command-line <code class="option">-D</code> option,
548 explained above, or by the meta commands explained below.
549 In addition to any variables preset by <code class="option">-D</code> command-line options,
550 there are a few variables that are preset automatically, listed in
551 <a class="xref" href="pgbench.html#PGBENCH-AUTOMATIC-VARIABLES" title="Table 301. pgbench Automatic Variables">Table 301</a>. A value specified for these
552 variables using <code class="option">-D</code> takes precedence over the automatic presets.
553 Once set, a variable's
554 value can be inserted into an SQL command by writing
555 <code class="literal">:</code><em class="replaceable"><code>variablename</code></em>. When running more than
556 one client session, each session has its own set of variables.
557 <span class="application">pgbench</span> supports up to 255 variable uses in one
559 </p><div class="table" id="PGBENCH-AUTOMATIC-VARIABLES"><p class="title"><strong>Table 301. pgbench Automatic Variables</strong></p><div class="table-contents"><table class="table" summary="pgbench Automatic Variables" border="1"><colgroup><col class="col1" /><col class="col2" /></colgroup><thead><tr><th>Variable</th><th>Description</th></tr></thead><tbody><tr><td> <code class="literal">client_id</code> </td><td>unique number identifying the client session (starts from zero)</td></tr><tr><td> <code class="literal">default_seed</code> </td><td>seed used in hash and pseudorandom permutation functions by default</td></tr><tr><td> <code class="literal">random_seed</code> </td><td>random generator seed (unless overwritten with <code class="option">-D</code>)</td></tr><tr><td> <code class="literal">scale</code> </td><td>current scale factor</td></tr></tbody></table></div></div><br class="table-break" /><p>
560 Script file meta commands begin with a backslash (<code class="literal">\</code>) and
561 normally extend to the end of the line, although they can be continued
562 to additional lines by writing backslash-return.
563 Arguments to a meta command are separated by white space.
564 These meta commands are supported:
565 </p><div class="variablelist"><dl class="variablelist"><dt id="PGBENCH-METACOMMAND-GSET"><span class="term">
566 <code class="literal">\gset [<em class="replaceable"><code>prefix</code></em>]</code>
567 <code class="literal">\aset [<em class="replaceable"><code>prefix</code></em>]</code>
568 </span> <a href="#PGBENCH-METACOMMAND-GSET" class="id_link">#</a></dt><dd><p>
569 These commands may be used to end SQL queries, taking the place of the
570 terminating semicolon (<code class="literal">;</code>).
572 When the <code class="literal">\gset</code> command is used, the preceding SQL query is
573 expected to return one row, the columns of which are stored into variables
574 named after column names, and prefixed with <em class="replaceable"><code>prefix</code></em>
577 When the <code class="literal">\aset</code> command is used, all combined SQL queries
578 (separated by <code class="literal">\;</code>) have their columns stored into variables
579 named after column names, and prefixed with <em class="replaceable"><code>prefix</code></em>
580 if provided. If a query returns no row, no assignment is made and the variable
581 can be tested for existence to detect this. If a query returns more than one
582 row, the last value is kept.
584 <code class="literal">\gset</code> and <code class="literal">\aset</code> cannot be used in
585 pipeline mode, since the query results are not yet available by the time
586 the commands would need them.
588 The following example puts the final account balance from the first query
589 into variable <em class="replaceable"><code>abalance</code></em>, and fills variables
590 <em class="replaceable"><code>p_two</code></em> and <em class="replaceable"><code>p_three</code></em>
591 with integers from the third query.
592 The result of the second query is discarded.
593 The result of the two last combined queries are stored in variables
594 <em class="replaceable"><code>four</code></em> and <em class="replaceable"><code>five</code></em>.
595 </p><pre class="programlisting">
596 UPDATE pgbench_accounts
597 SET abalance = abalance + :delta
599 RETURNING abalance \gset
600 -- compound of two queries
602 SELECT 2 AS two, 3 AS three \gset p_
603 SELECT 4 AS four \; SELECT 5 AS five \aset
604 </pre></dd><dt id="PGBENCH-METACOMMAND-IF-ELSE"><span class="term"><code class="literal">\if</code> <em class="replaceable"><code>expression</code></em><br /></span><span class="term"><code class="literal">\elif</code> <em class="replaceable"><code>expression</code></em><br /></span><span class="term"><code class="literal">\else</code><br /></span><span class="term"><code class="literal">\endif</code></span> <a href="#PGBENCH-METACOMMAND-IF-ELSE" class="id_link">#</a></dt><dd><p>
605 This group of commands implements nestable conditional blocks,
606 similarly to <code class="literal">psql</code>'s <a class="xref" href="app-psql.html#PSQL-METACOMMAND-IF"><code class="literal">\if</code> <em class="replaceable"><code>expression</code></em></a>.
607 Conditional expressions are identical to those with <code class="literal">\set</code>,
608 with non-zero values interpreted as true.
609 </p></dd><dt id="PGBENCH-METACOMMAND-SET"><span class="term">
610 <code class="literal">\set <em class="replaceable"><code>varname</code></em> <em class="replaceable"><code>expression</code></em></code>
611 </span> <a href="#PGBENCH-METACOMMAND-SET" class="id_link">#</a></dt><dd><p>
612 Sets variable <em class="replaceable"><code>varname</code></em> to a value calculated
613 from <em class="replaceable"><code>expression</code></em>.
614 The expression may contain the <code class="literal">NULL</code> constant,
615 Boolean constants <code class="literal">TRUE</code> and <code class="literal">FALSE</code>,
616 integer constants such as <code class="literal">5432</code>,
617 double constants such as <code class="literal">3.14159</code>,
618 references to variables <code class="literal">:</code><em class="replaceable"><code>variablename</code></em>,
619 <a class="link" href="pgbench.html#PGBENCH-BUILTIN-OPERATORS" title="Built-in Operators">operators</a>
620 with their usual SQL precedence and associativity,
621 <a class="link" href="pgbench.html#PGBENCH-BUILTIN-FUNCTIONS" title="Built-In Functions">function calls</a>,
622 SQL <a class="link" href="functions-conditional.html#FUNCTIONS-CASE" title="9.18.1. CASE"><code class="token">CASE</code> generic conditional
623 expressions</a> and parentheses.
625 Functions and most operators return <code class="literal">NULL</code> on
626 <code class="literal">NULL</code> input.
628 For conditional purposes, non zero numerical values are
629 <code class="literal">TRUE</code>, zero numerical values and <code class="literal">NULL</code>
630 are <code class="literal">FALSE</code>.
632 Too large or small integer and double constants, as well as
633 integer arithmetic operators (<code class="literal">+</code>,
634 <code class="literal">-</code>, <code class="literal">*</code> and <code class="literal">/</code>)
635 raise errors on overflows.
637 When no final <code class="token">ELSE</code> clause is provided to a
638 <code class="token">CASE</code>, the default value is <code class="literal">NULL</code>.
641 </p><pre class="programlisting">
642 \set ntellers 10 * :scale
643 \set aid (1021 * random(1, 100000 * :scale)) % \
644 (100000 * :scale) + 1
645 \set divx CASE WHEN :x <> 0 THEN :y/:x ELSE NULL END
646 </pre></dd><dt id="PGBENCH-METACOMMAND-SLEEP"><span class="term">
647 <code class="literal">\sleep <em class="replaceable"><code>number</code></em> [ us | ms | s ]</code>
648 </span> <a href="#PGBENCH-METACOMMAND-SLEEP" class="id_link">#</a></dt><dd><p>
649 Causes script execution to sleep for the specified duration in
650 microseconds (<code class="literal">us</code>), milliseconds (<code class="literal">ms</code>) or seconds
651 (<code class="literal">s</code>). If the unit is omitted then seconds are the default.
652 <em class="replaceable"><code>number</code></em> can be either an integer constant or a
653 <code class="literal">:</code><em class="replaceable"><code>variablename</code></em> reference to a variable
654 having an integer value.
657 </p><pre class="programlisting">
659 </pre></dd><dt id="PGBENCH-METACOMMAND-SETSHELL"><span class="term">
660 <code class="literal">\setshell <em class="replaceable"><code>varname</code></em> <em class="replaceable"><code>command</code></em> [ <em class="replaceable"><code>argument</code></em> ... ]</code>
661 </span> <a href="#PGBENCH-METACOMMAND-SETSHELL" class="id_link">#</a></dt><dd><p>
662 Sets variable <em class="replaceable"><code>varname</code></em> to the result of the shell command
663 <em class="replaceable"><code>command</code></em> with the given <em class="replaceable"><code>argument</code></em>(s).
664 The command must return an integer value through its standard output.
666 <em class="replaceable"><code>command</code></em> and each <em class="replaceable"><code>argument</code></em> can be either
667 a text constant or a <code class="literal">:</code><em class="replaceable"><code>variablename</code></em> reference
668 to a variable. If you want to use an <em class="replaceable"><code>argument</code></em> starting
669 with a colon, write an additional colon at the beginning of
670 <em class="replaceable"><code>argument</code></em>.
673 </p><pre class="programlisting">
674 \setshell variable_to_be_assigned command literal_argument :variable ::literal_starting_with_colon
675 </pre></dd><dt id="PGBENCH-METACOMMAND-SHELL"><span class="term">
676 <code class="literal">\shell <em class="replaceable"><code>command</code></em> [ <em class="replaceable"><code>argument</code></em> ... ]</code>
677 </span> <a href="#PGBENCH-METACOMMAND-SHELL" class="id_link">#</a></dt><dd><p>
678 Same as <code class="literal">\setshell</code>, but the result of the command
682 </p><pre class="programlisting">
683 \shell command literal_argument :variable ::literal_starting_with_colon
684 </pre></dd><dt id="PGBENCH-METACOMMAND-PIPELINE"><span class="term"><code class="literal">\startpipeline</code><br /></span><span class="term"><code class="literal">\syncpipeline</code><br /></span><span class="term"><code class="literal">\endpipeline</code></span> <a href="#PGBENCH-METACOMMAND-PIPELINE" class="id_link">#</a></dt><dd><p>
685 This group of commands implements pipelining of SQL statements.
686 A pipeline must begin with a <code class="command">\startpipeline</code>
687 and end with an <code class="command">\endpipeline</code>. In between there
688 may be any number of <code class="command">\syncpipeline</code> commands,
689 which sends a <a class="link" href="protocol-flow.html#PROTOCOL-FLOW-EXT-QUERY" title="54.2.3. Extended Query">sync message</a>
690 without ending the ongoing pipeline and flushing the send buffer.
691 In pipeline mode, statements are sent to the server without waiting
692 for the results of previous statements. See
693 <a class="xref" href="libpq-pipeline-mode.html" title="32.5. Pipeline Mode">Section 32.5</a> for more details.
694 Pipeline mode requires the use of extended query protocol.
695 </p></dd></dl></div></div><div class="refsect2" id="PGBENCH-BUILTIN-OPERATORS"><h3>Built-in Operators</h3><p>
696 The arithmetic, bitwise, comparison and logical operators listed in
697 <a class="xref" href="pgbench.html#PGBENCH-OPERATORS" title="Table 302. pgbench Operators">Table 302</a> are built into <span class="application">pgbench</span>
698 and may be used in expressions appearing in
699 <a class="link" href="pgbench.html#PGBENCH-METACOMMAND-SET"><code class="literal">\set</code></a>.
700 The operators are listed in increasing precedence order.
701 Except as noted, operators taking two numeric inputs will produce
702 a double value if either input is double, otherwise they produce
704 </p><div class="table" id="PGBENCH-OPERATORS"><p class="title"><strong>Table 302. pgbench Operators</strong></p><div class="table-contents"><table class="table" summary="pgbench Operators" border="1"><colgroup><col /></colgroup><thead><tr><th class="func_table_entry"><p class="func_signature">
712 </p></th></tr></thead><tbody><tr><td class="func_table_entry"><p class="func_signature">
713 <em class="replaceable"><code>boolean</code></em> <code class="literal">OR</code> <em class="replaceable"><code>boolean</code></em>
714 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
720 <code class="literal">5 or 0</code>
721 → <code class="returnvalue">TRUE</code>
722 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
723 <em class="replaceable"><code>boolean</code></em> <code class="literal">AND</code> <em class="replaceable"><code>boolean</code></em>
724 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
730 <code class="literal">3 and 0</code>
731 → <code class="returnvalue">FALSE</code>
732 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
733 <code class="literal">NOT</code> <em class="replaceable"><code>boolean</code></em>
734 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
740 <code class="literal">not false</code>
741 → <code class="returnvalue">TRUE</code>
742 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
743 <em class="replaceable"><code>boolean</code></em> <code class="literal">IS [NOT] (NULL|TRUE|FALSE)</code>
744 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
750 <code class="literal">1 is null</code>
751 → <code class="returnvalue">FALSE</code>
752 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
753 <em class="replaceable"><code>value</code></em> <code class="literal">ISNULL|NOTNULL</code>
754 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
760 <code class="literal">1 notnull</code>
761 → <code class="returnvalue">TRUE</code>
762 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
763 <em class="replaceable"><code>number</code></em> <code class="literal">=</code> <em class="replaceable"><code>number</code></em>
764 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
770 <code class="literal">5 = 4</code>
771 → <code class="returnvalue">FALSE</code>
772 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
773 <em class="replaceable"><code>number</code></em> <code class="literal"><></code> <em class="replaceable"><code>number</code></em>
774 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
780 <code class="literal">5 <> 4</code>
781 → <code class="returnvalue">TRUE</code>
782 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
783 <em class="replaceable"><code>number</code></em> <code class="literal">!=</code> <em class="replaceable"><code>number</code></em>
784 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
790 <code class="literal">5 != 5</code>
791 → <code class="returnvalue">FALSE</code>
792 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
793 <em class="replaceable"><code>number</code></em> <code class="literal"><</code> <em class="replaceable"><code>number</code></em>
794 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
800 <code class="literal">5 < 4</code>
801 → <code class="returnvalue">FALSE</code>
802 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
803 <em class="replaceable"><code>number</code></em> <code class="literal"><=</code> <em class="replaceable"><code>number</code></em>
804 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
807 Less than or equal to
810 <code class="literal">5 <= 4</code>
811 → <code class="returnvalue">FALSE</code>
812 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
813 <em class="replaceable"><code>number</code></em> <code class="literal">></code> <em class="replaceable"><code>number</code></em>
814 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
820 <code class="literal">5 > 4</code>
821 → <code class="returnvalue">TRUE</code>
822 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
823 <em class="replaceable"><code>number</code></em> <code class="literal">>=</code> <em class="replaceable"><code>number</code></em>
824 → <code class="returnvalue"><em class="replaceable"><code>boolean</code></em></code>
827 Greater than or equal to
830 <code class="literal">5 >= 4</code>
831 → <code class="returnvalue">TRUE</code>
832 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
833 <em class="replaceable"><code>integer</code></em> <code class="literal">|</code> <em class="replaceable"><code>integer</code></em>
834 → <code class="returnvalue"><em class="replaceable"><code>integer</code></em></code>
840 <code class="literal">1 | 2</code>
841 → <code class="returnvalue">3</code>
842 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
843 <em class="replaceable"><code>integer</code></em> <code class="literal">#</code> <em class="replaceable"><code>integer</code></em>
844 → <code class="returnvalue"><em class="replaceable"><code>integer</code></em></code>
850 <code class="literal">1 # 3</code>
851 → <code class="returnvalue">2</code>
852 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
853 <em class="replaceable"><code>integer</code></em> <code class="literal">&</code> <em class="replaceable"><code>integer</code></em>
854 → <code class="returnvalue"><em class="replaceable"><code>integer</code></em></code>
860 <code class="literal">1 & 3</code>
861 → <code class="returnvalue">1</code>
862 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
863 <code class="literal">~</code> <em class="replaceable"><code>integer</code></em>
864 → <code class="returnvalue"><em class="replaceable"><code>integer</code></em></code>
870 <code class="literal">~ 1</code>
871 → <code class="returnvalue">-2</code>
872 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
873 <em class="replaceable"><code>integer</code></em> <code class="literal"><<</code> <em class="replaceable"><code>integer</code></em>
874 → <code class="returnvalue"><em class="replaceable"><code>integer</code></em></code>
880 <code class="literal">1 << 2</code>
881 → <code class="returnvalue">4</code>
882 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
883 <em class="replaceable"><code>integer</code></em> <code class="literal">>></code> <em class="replaceable"><code>integer</code></em>
884 → <code class="returnvalue"><em class="replaceable"><code>integer</code></em></code>
890 <code class="literal">8 >> 2</code>
891 → <code class="returnvalue">2</code>
892 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
893 <em class="replaceable"><code>number</code></em> <code class="literal">+</code> <em class="replaceable"><code>number</code></em>
894 → <code class="returnvalue"><em class="replaceable"><code>number</code></em></code>
900 <code class="literal">5 + 4</code>
901 → <code class="returnvalue">9</code>
902 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
903 <em class="replaceable"><code>number</code></em> <code class="literal">-</code> <em class="replaceable"><code>number</code></em>
904 → <code class="returnvalue"><em class="replaceable"><code>number</code></em></code>
910 <code class="literal">3 - 2.0</code>
911 → <code class="returnvalue">1.0</code>
912 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
913 <em class="replaceable"><code>number</code></em> <code class="literal">*</code> <em class="replaceable"><code>number</code></em>
914 → <code class="returnvalue"><em class="replaceable"><code>number</code></em></code>
920 <code class="literal">5 * 4</code>
921 → <code class="returnvalue">20</code>
922 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
923 <em class="replaceable"><code>number</code></em> <code class="literal">/</code> <em class="replaceable"><code>number</code></em>
924 → <code class="returnvalue"><em class="replaceable"><code>number</code></em></code>
927 Division (truncates the result towards zero if both inputs are integers)
930 <code class="literal">5 / 3</code>
931 → <code class="returnvalue">1</code>
932 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
933 <em class="replaceable"><code>integer</code></em> <code class="literal">%</code> <em class="replaceable"><code>integer</code></em>
934 → <code class="returnvalue"><em class="replaceable"><code>integer</code></em></code>
940 <code class="literal">3 % 2</code>
941 → <code class="returnvalue">1</code>
942 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
943 <code class="literal">-</code> <em class="replaceable"><code>number</code></em>
944 → <code class="returnvalue"><em class="replaceable"><code>number</code></em></code>
950 <code class="literal">- 2.0</code>
951 → <code class="returnvalue">-2.0</code>
952 </p></td></tr></tbody></table></div></div><br class="table-break" /></div><div class="refsect2" id="PGBENCH-BUILTIN-FUNCTIONS"><h3>Built-In Functions</h3><p>
953 The functions listed in <a class="xref" href="pgbench.html#PGBENCH-FUNCTIONS" title="Table 303. pgbench Functions">Table 303</a> are built
954 into <span class="application">pgbench</span> and may be used in expressions appearing in
955 <a class="link" href="pgbench.html#PGBENCH-METACOMMAND-SET"><code class="literal">\set</code></a>.
956 </p><div class="table" id="PGBENCH-FUNCTIONS"><p class="title"><strong>Table 303. pgbench Functions</strong></p><div class="table-contents"><table class="table" summary="pgbench Functions" border="1"><colgroup><col /></colgroup><thead><tr><th class="func_table_entry"><p class="func_signature">
964 </p></th></tr></thead><tbody><tr><td class="func_table_entry"><p class="func_signature">
965 <code class="function">abs</code> ( <em class="replaceable"><code>number</code></em> )
966 → <code class="returnvalue"></code> same type as input
972 <code class="literal">abs(-17)</code>
973 → <code class="returnvalue">17</code>
974 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
975 <code class="function">debug</code> ( <em class="replaceable"><code>number</code></em> )
976 → <code class="returnvalue"></code> same type as input
979 Prints the argument to <span class="systemitem">stderr</span>,
980 and returns the argument.
983 <code class="literal">debug(5432.1)</code>
984 → <code class="returnvalue">5432.1</code>
985 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
986 <code class="function">double</code> ( <em class="replaceable"><code>number</code></em> )
987 → <code class="returnvalue">double</code>
993 <code class="literal">double(5432)</code>
994 → <code class="returnvalue">5432.0</code>
995 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
996 <code class="function">exp</code> ( <em class="replaceable"><code>number</code></em> )
997 → <code class="returnvalue">double</code>
1000 Exponential (<code class="literal">e</code> raised to the given power)
1003 <code class="literal">exp(1.0)</code>
1004 → <code class="returnvalue">2.718281828459045</code>
1005 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1006 <code class="function">greatest</code> ( <em class="replaceable"><code>number</code></em> [<span class="optional">, <code class="literal">...</code> </span>] )
1007 → <code class="returnvalue"></code> <code class="type">double</code> if any argument is double, else <code class="type">integer</code>
1010 Selects the largest value among the arguments.
1013 <code class="literal">greatest(5, 4, 3, 2)</code>
1014 → <code class="returnvalue">5</code>
1015 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1016 <code class="function">hash</code> ( <em class="parameter"><code>value</code></em> [<span class="optional">, <em class="parameter"><code>seed</code></em> </span>] )
1017 → <code class="returnvalue">integer</code>
1020 This is an alias for <code class="function">hash_murmur2</code>.
1023 <code class="literal">hash(10, 5432)</code>
1024 → <code class="returnvalue">-5817877081768721676</code>
1025 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1026 <code class="function">hash_fnv1a</code> ( <em class="parameter"><code>value</code></em> [<span class="optional">, <em class="parameter"><code>seed</code></em> </span>] )
1027 → <code class="returnvalue">integer</code>
1030 Computes <a class="ulink" href="https://en.wikipedia.org/wiki/Fowler%E2%80%93Noll%E2%80%93Vo_hash_function" target="_top">FNV-1a hash</a>.
1033 <code class="literal">hash_fnv1a(10, 5432)</code>
1034 → <code class="returnvalue">-7793829335365542153</code>
1035 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1036 <code class="function">hash_murmur2</code> ( <em class="parameter"><code>value</code></em> [<span class="optional">, <em class="parameter"><code>seed</code></em> </span>] )
1037 → <code class="returnvalue">integer</code>
1040 Computes <a class="ulink" href="https://en.wikipedia.org/wiki/MurmurHash" target="_top">MurmurHash2 hash</a>.
1043 <code class="literal">hash_murmur2(10, 5432)</code>
1044 → <code class="returnvalue">-5817877081768721676</code>
1045 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1046 <code class="function">int</code> ( <em class="replaceable"><code>number</code></em> )
1047 → <code class="returnvalue">integer</code>
1053 <code class="literal">int(5.4 + 3.8)</code>
1054 → <code class="returnvalue">9</code>
1055 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1056 <code class="function">least</code> ( <em class="replaceable"><code>number</code></em> [<span class="optional">, <code class="literal">...</code> </span>] )
1057 → <code class="returnvalue"></code> <code class="type">double</code> if any argument is double, else <code class="type">integer</code>
1060 Selects the smallest value among the arguments.
1063 <code class="literal">least(5, 4, 3, 2.1)</code>
1064 → <code class="returnvalue">2.1</code>
1065 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1066 <code class="function">ln</code> ( <em class="replaceable"><code>number</code></em> )
1067 → <code class="returnvalue">double</code>
1073 <code class="literal">ln(2.718281828459045)</code>
1074 → <code class="returnvalue">1.0</code>
1075 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1076 <code class="function">mod</code> ( <em class="replaceable"><code>integer</code></em>, <em class="replaceable"><code>integer</code></em> )
1077 → <code class="returnvalue">integer</code>
1083 <code class="literal">mod(54, 32)</code>
1084 → <code class="returnvalue">22</code>
1085 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1086 <code class="function">permute</code> ( <em class="parameter"><code>i</code></em>, <em class="parameter"><code>size</code></em> [, <em class="parameter"><code>seed</code></em> ] )
1087 → <code class="returnvalue">integer</code>
1090 Permuted value of <em class="parameter"><code>i</code></em>, in the range
1091 <code class="literal">[0, size)</code>. This is the new position of
1092 <em class="parameter"><code>i</code></em> (modulo <em class="parameter"><code>size</code></em>) in a
1093 pseudorandom permutation of the integers <code class="literal">0...size-1</code>,
1094 parameterized by <em class="parameter"><code>seed</code></em>, see below.
1097 <code class="literal">permute(0, 4)</code>
1098 → <code class="returnvalue">an integer between 0 and 3</code>
1099 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1100 <code class="function">pi</code> ()
1101 → <code class="returnvalue">double</code>
1104 Approximate value of <span class="symbol_font">π</span>
1107 <code class="literal">pi()</code>
1108 → <code class="returnvalue">3.14159265358979323846</code>
1109 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1110 <code class="function">pow</code> ( <em class="parameter"><code>x</code></em>, <em class="parameter"><code>y</code></em> )
1111 → <code class="returnvalue">double</code>
1113 <p class="func_signature">
1114 <code class="function">power</code> ( <em class="parameter"><code>x</code></em>, <em class="parameter"><code>y</code></em> )
1115 → <code class="returnvalue">double</code>
1118 <em class="parameter"><code>x</code></em> raised to the power of <em class="parameter"><code>y</code></em>
1121 <code class="literal">pow(2.0, 10)</code>
1122 → <code class="returnvalue">1024.0</code>
1123 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1124 <code class="function">random</code> ( <em class="parameter"><code>lb</code></em>, <em class="parameter"><code>ub</code></em> )
1125 → <code class="returnvalue">integer</code>
1128 Computes a uniformly-distributed random integer in <code class="literal">[lb,
1132 <code class="literal">random(1, 10)</code>
1133 → <code class="returnvalue">an integer between 1 and 10</code>
1134 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1135 <code class="function">random_exponential</code> ( <em class="parameter"><code>lb</code></em>, <em class="parameter"><code>ub</code></em>, <em class="parameter"><code>parameter</code></em> )
1136 → <code class="returnvalue">integer</code>
1139 Computes an exponentially-distributed random integer in <code class="literal">[lb,
1140 ub]</code>, see below.
1143 <code class="literal">random_exponential(1, 10, 3.0)</code>
1144 → <code class="returnvalue">an integer between 1 and 10</code>
1145 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1146 <code class="function">random_gaussian</code> ( <em class="parameter"><code>lb</code></em>, <em class="parameter"><code>ub</code></em>, <em class="parameter"><code>parameter</code></em> )
1147 → <code class="returnvalue">integer</code>
1150 Computes a Gaussian-distributed random integer in <code class="literal">[lb,
1151 ub]</code>, see below.
1154 <code class="literal">random_gaussian(1, 10, 2.5)</code>
1155 → <code class="returnvalue">an integer between 1 and 10</code>
1156 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1157 <code class="function">random_zipfian</code> ( <em class="parameter"><code>lb</code></em>, <em class="parameter"><code>ub</code></em>, <em class="parameter"><code>parameter</code></em> )
1158 → <code class="returnvalue">integer</code>
1161 Computes a Zipfian-distributed random integer in <code class="literal">[lb,
1162 ub]</code>, see below.
1165 <code class="literal">random_zipfian(1, 10, 1.5)</code>
1166 → <code class="returnvalue">an integer between 1 and 10</code>
1167 </p></td></tr><tr><td class="func_table_entry"><p class="func_signature">
1168 <code class="function">sqrt</code> ( <em class="replaceable"><code>number</code></em> )
1169 → <code class="returnvalue">double</code>
1175 <code class="literal">sqrt(2.0)</code>
1176 → <code class="returnvalue">1.414213562</code>
1177 </p></td></tr></tbody></table></div></div><br class="table-break" /><p>
1178 The <code class="literal">random</code> function generates values using a uniform
1179 distribution, that is all the values are drawn within the specified
1180 range with equal probability. The <code class="literal">random_exponential</code>,
1181 <code class="literal">random_gaussian</code> and <code class="literal">random_zipfian</code>
1182 functions require an additional double parameter which determines the precise
1183 shape of the distribution.
1184 </p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
1185 For an exponential distribution, <em class="replaceable"><code>parameter</code></em>
1186 controls the distribution by truncating a quickly-decreasing
1187 exponential distribution at <em class="replaceable"><code>parameter</code></em>, and then
1188 projecting onto integers between the bounds.
1190 </p><div class="literallayout"><p><br />
1191 f(x) = exp(-parameter * (x - min) / (max - min + 1)) / (1 - exp(-parameter))<br />
1193 Then value <em class="replaceable"><code>i</code></em> between <em class="replaceable"><code>min</code></em> and
1194 <em class="replaceable"><code>max</code></em> inclusive is drawn with probability:
1195 <code class="literal">f(i) - f(i + 1)</code>.
1197 Intuitively, the larger the <em class="replaceable"><code>parameter</code></em>, the more
1198 frequently values close to <em class="replaceable"><code>min</code></em> are accessed, and the
1199 less frequently values close to <em class="replaceable"><code>max</code></em> are accessed.
1200 The closer to 0 <em class="replaceable"><code>parameter</code></em> is, the flatter (more
1201 uniform) the access distribution.
1202 A crude approximation of the distribution is that the most frequent 1%
1203 values in the range, close to <em class="replaceable"><code>min</code></em>, are drawn
1204 <em class="replaceable"><code>parameter</code></em>% of the time.
1205 The <em class="replaceable"><code>parameter</code></em> value must be strictly positive.
1206 </p></li><li class="listitem"><p>
1207 For a Gaussian distribution, the interval is mapped onto a standard
1208 normal distribution (the classical bell-shaped Gaussian curve) truncated
1209 at <code class="literal">-parameter</code> on the left and <code class="literal">+parameter</code>
1211 Values in the middle of the interval are more likely to be drawn.
1212 To be precise, if <code class="literal">PHI(x)</code> is the cumulative distribution
1213 function of the standard normal distribution, with mean <code class="literal">mu</code>
1214 defined as <code class="literal">(max + min) / 2.0</code>, with
1215 </p><div class="literallayout"><p><br />
1216 f(x) = PHI(2.0 * parameter * (x - mu) / (max - min + 1)) /<br />
1217 (2.0 * PHI(parameter) - 1)<br />
1219 then value <em class="replaceable"><code>i</code></em> between <em class="replaceable"><code>min</code></em> and
1220 <em class="replaceable"><code>max</code></em> inclusive is drawn with probability:
1221 <code class="literal">f(i + 0.5) - f(i - 0.5)</code>.
1222 Intuitively, the larger the <em class="replaceable"><code>parameter</code></em>, the more
1223 frequently values close to the middle of the interval are drawn, and the
1224 less frequently values close to the <em class="replaceable"><code>min</code></em> and
1225 <em class="replaceable"><code>max</code></em> bounds. About 67% of values are drawn from the
1226 middle <code class="literal">1.0 / parameter</code>, that is a relative
1227 <code class="literal">0.5 / parameter</code> around the mean, and 95% in the middle
1228 <code class="literal">2.0 / parameter</code>, that is a relative
1229 <code class="literal">1.0 / parameter</code> around the mean; for instance, if
1230 <em class="replaceable"><code>parameter</code></em> is 4.0, 67% of values are drawn from the
1231 middle quarter (1.0 / 4.0) of the interval (i.e., from
1232 <code class="literal">3.0 / 8.0</code> to <code class="literal">5.0 / 8.0</code>) and 95% from
1233 the middle half (<code class="literal">2.0 / 4.0</code>) of the interval (second and third
1234 quartiles). The minimum allowed <em class="replaceable"><code>parameter</code></em>
1236 </p></li><li class="listitem"><p>
1237 <code class="literal">random_zipfian</code> generates a bounded Zipfian
1239 <em class="replaceable"><code>parameter</code></em> defines how skewed the distribution
1240 is. The larger the <em class="replaceable"><code>parameter</code></em>, the more
1241 frequently values closer to the beginning of the interval are drawn.
1242 The distribution is such that, assuming the range starts from 1,
1243 the ratio of the probability of drawing <em class="replaceable"><code>k</code></em>
1244 versus drawing <em class="replaceable"><code>k+1</code></em> is
1245 <code class="literal">((<em class="replaceable"><code>k</code></em>+1)/<em class="replaceable"><code>k</code></em>)**<em class="replaceable"><code>parameter</code></em></code>.
1246 For example, <code class="literal">random_zipfian(1, ..., 2.5)</code> produces
1247 the value <code class="literal">1</code> about <code class="literal">(2/1)**2.5 =
1248 5.66</code> times more frequently than <code class="literal">2</code>, which
1249 itself is produced <code class="literal">(3/2)**2.5 = 2.76</code> times more
1250 frequently than <code class="literal">3</code>, and so on.
1252 <span class="application">pgbench</span>'s implementation is based on
1253 "Non-Uniform Random Variate Generation", Luc Devroye, p. 550-551,
1254 Springer 1986. Due to limitations of that algorithm,
1255 the <em class="replaceable"><code>parameter</code></em> value is restricted to
1256 the range [1.001, 1000].
1257 </p></li></ul></div><div class="note"><h3 class="title">Note</h3><p>
1258 When designing a benchmark which selects rows non-uniformly, be aware
1259 that the rows chosen may be correlated with other data such as IDs from
1260 a sequence or the physical row ordering, which may skew performance
1263 To avoid this, you may wish to use the <code class="function">permute</code>
1264 function, or some other additional step with similar effect, to shuffle
1265 the selected rows and remove such correlations.
1267 Hash functions <code class="literal">hash</code>, <code class="literal">hash_murmur2</code> and
1268 <code class="literal">hash_fnv1a</code> accept an input value and an optional seed parameter.
1269 In case the seed isn't provided the value of <code class="literal">:default_seed</code>
1270 is used, which is initialized randomly unless set by the command-line
1271 <code class="literal">-D</code> option.
1273 <code class="literal">permute</code> accepts an input value, a size, and an optional
1274 seed parameter. It generates a pseudorandom permutation of integers in
1275 the range <code class="literal">[0, size)</code>, and returns the index of the input
1276 value in the permuted values. The permutation chosen is parameterized by
1277 the seed, which defaults to <code class="literal">:default_seed</code>, if not
1278 specified. Unlike the hash functions, <code class="literal">permute</code> ensures
1279 that there are no collisions or holes in the output values. Input values
1280 outside the interval are interpreted modulo the size. The function raises
1281 an error if the size is not positive. <code class="function">permute</code> can be
1282 used to scatter the distribution of non-uniform random functions such as
1283 <code class="literal">random_zipfian</code> or <code class="literal">random_exponential</code>
1284 so that values drawn more often are not trivially correlated. For
1285 instance, the following <span class="application">pgbench</span> script
1286 simulates a possible real world workload typical for social media and
1287 blogging platforms where a few accounts generate excessive load:
1289 </p><pre class="programlisting">
1291 \set r random_zipfian(1, :size, 1.07)
1292 \set k 1 + permute(:r, :size)
1295 In some cases several distinct distributions are needed which don't correlate
1296 with each other and this is when the optional seed parameter comes in handy:
1298 </p><pre class="programlisting">
1299 \set k1 1 + permute(:r, :size, :default_seed + 123)
1300 \set k2 1 + permute(:r, :size, :default_seed + 321)
1303 A similar behavior can also be approximated with <code class="function">hash</code>:
1305 </p><pre class="programlisting">
1307 \set r random_zipfian(1, 100 * :size, 1.07)
1308 \set k 1 + abs(hash(:r)) % :size
1311 However, since <code class="function">hash</code> generates collisions, some values
1312 will not be reachable and others will be more frequent than expected from
1313 the original distribution.
1315 As an example, the full definition of the built-in TPC-B-like
1318 </p><pre class="programlisting">
1319 \set aid random(1, 100000 * :scale)
1320 \set bid random(1, 1 * :scale)
1321 \set tid random(1, 10 * :scale)
1322 \set delta random(-5000, 5000)
1324 UPDATE pgbench_accounts SET abalance = abalance + :delta WHERE aid = :aid;
1325 SELECT abalance FROM pgbench_accounts WHERE aid = :aid;
1326 UPDATE pgbench_tellers SET tbalance = tbalance + :delta WHERE tid = :tid;
1327 UPDATE pgbench_branches SET bbalance = bbalance + :delta WHERE bid = :bid;
1328 INSERT INTO pgbench_history (tid, bid, aid, delta, mtime) VALUES (:tid, :bid, :aid, :delta, CURRENT_TIMESTAMP);
1332 This script allows each iteration of the transaction to reference
1333 different, randomly-chosen rows. (This example also shows why it's
1334 important for each client session to have its own variables —
1335 otherwise they'd not be independently touching different rows.)
1336 </p></div><div class="refsect2" id="id-1.9.4.11.9.6"><h3>Per-Transaction Logging</h3><p>
1337 With the <code class="option">-l</code> option (but without
1338 the <code class="option">--aggregate-interval</code> option),
1339 <span class="application">pgbench</span> writes information about each transaction
1340 to a log file. The log file will be named
1341 <code class="filename"><em class="replaceable"><code>prefix</code></em>.<em class="replaceable"><code>nnn</code></em></code>,
1342 where <em class="replaceable"><code>prefix</code></em> defaults to <code class="literal">pgbench_log</code>, and
1343 <em class="replaceable"><code>nnn</code></em> is the PID of the
1344 <span class="application">pgbench</span> process.
1345 The prefix can be changed by using the <code class="option">--log-prefix</code> option.
1346 If the <code class="option">-j</code> option is 2 or higher, so that there are multiple
1347 worker threads, each will have its own log file. The first worker will
1348 use the same name for its log file as in the standard single worker case.
1349 The additional log files for the other workers will be named
1350 <code class="filename"><em class="replaceable"><code>prefix</code></em>.<em class="replaceable"><code>nnn</code></em>.<em class="replaceable"><code>mmm</code></em></code>,
1351 where <em class="replaceable"><code>mmm</code></em> is a sequential number for each worker starting
1354 Each line in a log file describes one transaction.
1355 It contains the following space-separated fields:
1357 </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><em class="replaceable"><code>client_id</code></em></span></dt><dd><p>
1358 identifies the client session that ran the transaction
1359 </p></dd><dt><span class="term"><em class="replaceable"><code>transaction_no</code></em></span></dt><dd><p>
1360 counts how many transactions have been run by that session
1361 </p></dd><dt><span class="term"><em class="replaceable"><code>time</code></em></span></dt><dd><p>
1362 transaction's elapsed time, in microseconds
1363 </p></dd><dt><span class="term"><em class="replaceable"><code>script_no</code></em></span></dt><dd><p>
1364 identifies the script file that was used for the transaction
1365 (useful when multiple scripts are specified
1366 with <code class="option">-f</code> or <code class="option">-b</code>)
1367 </p></dd><dt><span class="term"><em class="replaceable"><code>time_epoch</code></em></span></dt><dd><p>
1368 transaction's completion time, as a Unix-epoch time stamp
1369 </p></dd><dt><span class="term"><em class="replaceable"><code>time_us</code></em></span></dt><dd><p>
1370 fractional-second part of transaction's completion time, in
1372 </p></dd><dt><span class="term"><em class="replaceable"><code>schedule_lag</code></em></span></dt><dd><p>
1373 transaction start delay, that is the difference between the
1374 transaction's scheduled start time and the time it actually
1375 started, in microseconds
1376 (present only if <code class="option">--rate</code> is specified)
1377 </p></dd><dt><span class="term"><em class="replaceable"><code>retries</code></em></span></dt><dd><p>
1378 count of retries after serialization or deadlock errors during the
1380 (present only if <code class="option">--max-tries</code> is not equal to one)
1381 </p></dd></dl></div><p>
1383 When both <code class="option">--rate</code> and <code class="option">--latency-limit</code> are used,
1384 the <em class="replaceable"><code>time</code></em> for a skipped transaction will be reported as
1385 <code class="literal">skipped</code>.
1386 If the transaction ends with a failure, its <em class="replaceable"><code>time</code></em>
1387 will be reported as <code class="literal">failed</code>. If you use the
1388 <code class="option">--failures-detailed</code> option, the
1389 <em class="replaceable"><code>time</code></em> of the failed transaction will be reported as
1390 <code class="literal">serialization</code> or
1391 <code class="literal">deadlock</code> depending on the type of failure (see
1392 <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a> for more information).
1394 Here is a snippet of a log file generated in a single-client run:
1395 </p><pre class="screen">
1396 0 199 2241 0 1175850568 995598
1397 0 200 2465 0 1175850568 998079
1398 0 201 2513 0 1175850569 608
1399 0 202 2038 0 1175850569 2663
1402 Another example with <code class="literal">--rate=100</code>
1403 and <code class="literal">--latency-limit=5</code> (note the additional
1404 <em class="replaceable"><code>schedule_lag</code></em> column):
1405 </p><pre class="screen">
1406 0 81 4621 0 1412881037 912698 3005
1407 0 82 6173 0 1412881037 914578 4304
1408 0 83 skipped 0 1412881037 914578 5217
1409 0 83 skipped 0 1412881037 914578 5099
1410 0 83 4722 0 1412881037 916203 3108
1411 0 84 4142 0 1412881037 918023 2333
1412 0 85 2465 0 1412881037 919759 740
1414 In this example, transaction 82 was late, because its latency (6.173 ms) was
1415 over the 5 ms limit. The next two transactions were skipped, because they
1416 were already late before they were even started.
1418 The following example shows a snippet of a log file with failures and
1419 retries, with the maximum number of tries set to 10 (note the additional
1420 <em class="replaceable"><code>retries</code></em> column):
1421 </p><pre class="screen">
1422 3 0 47423 0 1499414498 34501 3
1423 3 1 8333 0 1499414498 42848 0
1424 3 2 8358 0 1499414498 51219 0
1425 4 0 72345 0 1499414498 59433 6
1426 1 3 41718 0 1499414498 67879 4
1427 1 4 8416 0 1499414498 76311 0
1428 3 3 33235 0 1499414498 84469 3
1429 0 0 failed 0 1499414498 84905 9
1430 2 0 failed 0 1499414498 86248 9
1431 3 4 8307 0 1499414498 92788 0
1434 If the <code class="option">--failures-detailed</code> option is used, the type of
1435 failure is reported in the <em class="replaceable"><code>time</code></em> like this:
1436 </p><pre class="screen">
1437 3 0 47423 0 1499414498 34501 3
1438 3 1 8333 0 1499414498 42848 0
1439 3 2 8358 0 1499414498 51219 0
1440 4 0 72345 0 1499414498 59433 6
1441 1 3 41718 0 1499414498 67879 4
1442 1 4 8416 0 1499414498 76311 0
1443 3 3 33235 0 1499414498 84469 3
1444 0 0 serialization 0 1499414498 84905 9
1445 2 0 serialization 0 1499414498 86248 9
1446 3 4 8307 0 1499414498 92788 0
1449 When running a long test on hardware that can handle a lot of transactions,
1450 the log files can become very large. The <code class="option">--sampling-rate</code> option
1451 can be used to log only a random sample of transactions.
1452 </p></div><div class="refsect2" id="id-1.9.4.11.9.7"><h3>Aggregated Logging</h3><p>
1453 With the <code class="option">--aggregate-interval</code> option, a different
1454 format is used for the log files. Each log line describes one
1455 aggregation interval. It contains the following space-separated
1458 </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><em class="replaceable"><code>interval_start</code></em></span></dt><dd><p>
1459 start time of the interval, as a Unix-epoch time stamp
1460 </p></dd><dt><span class="term"><em class="replaceable"><code>num_transactions</code></em></span></dt><dd><p>
1461 number of transactions within the interval
1462 </p></dd><dt><span class="term"><em class="replaceable"><code>sum_latency</code></em></span></dt><dd><p>
1463 sum of transaction latencies
1464 </p></dd><dt><span class="term"><em class="replaceable"><code>sum_latency_2</code></em></span></dt><dd><p>
1465 sum of squares of transaction latencies
1466 </p></dd><dt><span class="term"><em class="replaceable"><code>min_latency</code></em></span></dt><dd><p>
1467 minimum transaction latency
1468 </p></dd><dt><span class="term"><em class="replaceable"><code>max_latency</code></em></span></dt><dd><p>
1469 maximum transaction latency
1470 </p></dd><dt><span class="term"><em class="replaceable"><code>sum_lag</code></em></span></dt><dd><p>
1471 sum of transaction start delays
1472 (zero unless <code class="option">--rate</code> is specified)
1473 </p></dd><dt><span class="term"><em class="replaceable"><code>sum_lag_2</code></em></span></dt><dd><p>
1474 sum of squares of transaction start delays
1475 (zero unless <code class="option">--rate</code> is specified)
1476 </p></dd><dt><span class="term"><em class="replaceable"><code>min_lag</code></em></span></dt><dd><p>
1477 minimum transaction start delay
1478 (zero unless <code class="option">--rate</code> is specified)
1479 </p></dd><dt><span class="term"><em class="replaceable"><code>max_lag</code></em></span></dt><dd><p>
1480 maximum transaction start delay
1481 (zero unless <code class="option">--rate</code> is specified)
1482 </p></dd><dt><span class="term"><em class="replaceable"><code>skipped</code></em></span></dt><dd><p>
1483 number of transactions skipped because they would have started too late
1484 (zero unless <code class="option">--rate</code>
1485 and <code class="option">--latency-limit</code> are specified)
1486 </p></dd><dt><span class="term"><em class="replaceable"><code>retried</code></em></span></dt><dd><p>
1487 number of retried transactions
1488 (zero unless <code class="option">--max-tries</code> is not equal to one)
1489 </p></dd><dt><span class="term"><em class="replaceable"><code>retries</code></em></span></dt><dd><p>
1490 number of retries after serialization or deadlock errors
1491 (zero unless <code class="option">--max-tries</code> is not equal to one)
1492 </p></dd><dt><span class="term"><em class="replaceable"><code>serialization_failures</code></em></span></dt><dd><p>
1493 number of transactions that got a serialization error and were not
1495 (zero unless <code class="option">--failures-detailed</code> is specified)
1496 </p></dd><dt><span class="term"><em class="replaceable"><code>deadlock_failures</code></em></span></dt><dd><p>
1497 number of transactions that got a deadlock error and were not
1499 (zero unless <code class="option">--failures-detailed</code> is specified)
1500 </p></dd></dl></div><p>
1502 Here is some example output generated with this option:
1503 </p><pre class="screen">
1504 <strong class="userinput"><code>pgbench --aggregate-interval=10 --time=20 --client=10 --log --rate=1000 --latency-limit=10 --failures-detailed --max-tries=10 test</code></strong>
1506 1650260552 5178 26171317 177284491527 1136 44462 2647617 7321113867 0 9866 64 7564 28340 4148 0
1507 1650260562 4808 25573984 220121792172 1171 62083 3037380 9666800914 0 9998 598 7392 26621 4527 0
1510 Notice that while the plain (unaggregated) log format shows which script
1511 was used for each transaction, the aggregated format does not. Therefore if
1512 you need per-script data, you need to aggregate the data on your own.
1513 </p></div><div class="refsect2" id="id-1.9.4.11.9.8"><h3>Per-Statement Report</h3><p>
1514 With the <code class="option">-r</code> option, <span class="application">pgbench</span>
1515 collects the following statistics for each statement:
1516 </p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
1517 <code class="literal">latency</code> — elapsed transaction time for each
1518 statement. <span class="application">pgbench</span> reports an average value
1519 of all successful runs of the statement.
1520 </p></li><li class="listitem"><p>
1521 The number of failures in this statement. See
1522 <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a> for more information.
1523 </p></li><li class="listitem"><p>
1524 The number of retries after a serialization or a deadlock error in this
1525 statement. See <a class="xref" href="pgbench.html#FAILURES-AND-RETRIES" title="Failures and Serialization/Deadlock Retries">Failures and Serialization/Deadlock Retries</a> for more information.
1526 </p></li></ul></div><p>
1528 The report displays retry statistics only if the <code class="option">--max-tries</code>
1529 option is not equal to 1.
1531 All values are computed for each statement executed by every client and are
1532 reported after the benchmark has finished.
1534 For the default script, the output will look similar to this:
1535 </p><pre class="screen">
1536 starting vacuum...end.
1537 transaction type: <builtin: TPC-B (sort of)>
1540 number of clients: 10
1541 number of threads: 1
1542 maximum number of tries: 1
1543 number of transactions per client: 1000
1544 number of transactions actually processed: 10000/10000
1545 number of failed transactions: 0 (0.000%)
1546 number of transactions above the 50.0 ms latency limit: 1311/10000 (13.110 %)
1547 latency average = 28.488 ms
1548 latency stddev = 21.009 ms
1549 initial connection time = 69.068 ms
1550 tps = 346.224794 (without initial connection time)
1551 statement latencies in milliseconds and failures:
1552 0.012 0 \set aid random(1, 100000 * :scale)
1553 0.002 0 \set bid random(1, 1 * :scale)
1554 0.002 0 \set tid random(1, 10 * :scale)
1555 0.002 0 \set delta random(-5000, 5000)
1557 0.834 0 UPDATE pgbench_accounts SET abalance = abalance + :delta WHERE aid = :aid;
1558 0.641 0 SELECT abalance FROM pgbench_accounts WHERE aid = :aid;
1559 11.126 0 UPDATE pgbench_tellers SET tbalance = tbalance + :delta WHERE tid = :tid;
1560 12.961 0 UPDATE pgbench_branches SET bbalance = bbalance + :delta WHERE bid = :bid;
1561 0.634 0 INSERT INTO pgbench_history (tid, bid, aid, delta, mtime) VALUES (:tid, :bid, :aid, :delta, CURRENT_TIMESTAMP);
1565 Another example of output for the default script using serializable default
1566 transaction isolation level (<code class="command">PGOPTIONS='-c
1567 default_transaction_isolation=serializable' pgbench ...</code>):
1568 </p><pre class="screen">
1569 starting vacuum...end.
1570 transaction type: <builtin: TPC-B (sort of)>
1573 number of clients: 10
1574 number of threads: 1
1575 maximum number of tries: 10
1576 number of transactions per client: 1000
1577 number of transactions actually processed: 6317/10000
1578 number of failed transactions: 3683 (36.830%)
1579 number of transactions retried: 7667 (76.670%)
1580 total number of retries: 45339
1581 number of transactions above the 50.0 ms latency limit: 106/6317 (1.678 %)
1582 latency average = 17.016 ms
1583 latency stddev = 13.283 ms
1584 initial connection time = 45.017 ms
1585 tps = 186.792667 (without initial connection time)
1586 statement latencies in milliseconds, failures and retries:
1587 0.006 0 0 \set aid random(1, 100000 * :scale)
1588 0.001 0 0 \set bid random(1, 1 * :scale)
1589 0.001 0 0 \set tid random(1, 10 * :scale)
1590 0.001 0 0 \set delta random(-5000, 5000)
1592 0.773 0 1 UPDATE pgbench_accounts SET abalance = abalance + :delta WHERE aid = :aid;
1593 0.624 0 0 SELECT abalance FROM pgbench_accounts WHERE aid = :aid;
1594 1.098 320 3762 UPDATE pgbench_tellers SET tbalance = tbalance + :delta WHERE tid = :tid;
1595 0.582 3363 41576 UPDATE pgbench_branches SET bbalance = bbalance + :delta WHERE bid = :bid;
1596 0.465 0 0 INSERT INTO pgbench_history (tid, bid, aid, delta, mtime) VALUES (:tid, :bid, :aid, :delta, CURRENT_TIMESTAMP);
1599 If multiple script files are specified, all statistics are reported
1600 separately for each script file.
1602 Note that collecting the additional timing information needed for
1603 per-statement latency computation adds some overhead. This will slow
1604 average execution speed and lower the computed TPS. The amount
1605 of slowdown varies significantly depending on platform and hardware.
1606 Comparing average TPS values with and without latency reporting enabled
1607 is a good way to measure if the timing overhead is significant.
1608 </p></div><div class="refsect2" id="FAILURES-AND-RETRIES"><h3>Failures and Serialization/Deadlock Retries</h3><p>
1609 When executing <span class="application">pgbench</span>, there are three main types
1611 </p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
1612 Errors of the main program. They are the most serious and always result
1613 in an immediate exit from <span class="application">pgbench</span> with the
1614 corresponding error message. They include:
1615 </p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: circle; "><li class="listitem"><p>
1616 errors at the beginning of <span class="application">pgbench</span>
1617 (e.g. an invalid option value);
1618 </p></li><li class="listitem"><p>
1619 errors in the initialization mode (e.g. the query to create
1620 tables for built-in scripts fails);
1621 </p></li><li class="listitem"><p>
1622 errors before starting threads (e.g. could not connect to the
1623 database server, syntax error in the meta command, thread
1625 </p></li><li class="listitem"><p>
1626 internal <span class="application">pgbench</span> errors (which are
1627 supposed to never occur...).
1628 </p></li></ul></div></li><li class="listitem"><p>
1629 Errors when the thread manages its clients (e.g. the client could not
1630 start a connection to the database server / the socket for connecting
1631 the client to the database server has become invalid). In such cases
1632 all clients of this thread stop while other threads continue to work.
1633 However, <code class="option">--exit-on-abort</code> is specified, all of the
1634 threads stop immediately in this case.
1635 </p></li><li class="listitem"><p>
1636 Direct client errors. They lead to immediate exit from
1637 <span class="application">pgbench</span> with the corresponding error message
1638 in the case of an internal <span class="application">pgbench</span>
1639 error (which are supposed to never occur...) or when
1640 <code class="option">--exit-on-abort</code> is specified. Otherwise in the worst
1641 case they only lead to the abortion of the failed client while other
1642 clients continue their run (but some client errors are handled without
1643 an abortion of the client and reported separately, see below). Later in
1644 this section it is assumed that the discussed errors are only the
1645 direct client errors and they are not internal
1646 <span class="application">pgbench</span> errors.
1647 </p></li></ul></div><p>
1649 A client's run is aborted in case of a serious error; for example, the
1650 connection with the database server was lost or the end of script was reached
1651 without completing the last transaction. In addition, if execution of an SQL
1652 or meta command fails for reasons other than serialization or deadlock errors,
1653 the client is aborted. Otherwise, if an SQL command fails with serialization or
1654 deadlock errors, the client is not aborted. In such cases, the current
1655 transaction is rolled back, which also includes setting the client variables
1656 as they were before the run of this transaction (it is assumed that one
1657 transaction script contains only one transaction; see
1658 <a class="xref" href="pgbench.html#TRANSACTIONS-AND-SCRIPTS" title="What Is the “Transaction” Actually Performed in pgbench?">What Is the "Transaction" Actually Performed in pgbench?</a> for more information).
1659 Transactions with serialization or deadlock errors are repeated after
1660 rollbacks until they complete successfully or reach the maximum
1661 number of tries (specified by the <code class="option">--max-tries</code> option) / the maximum
1662 time of retries (specified by the <code class="option">--latency-limit</code> option) / the end
1663 of benchmark (specified by the <code class="option">--time</code> option). If
1664 the last trial run fails, this transaction will be reported as failed but
1665 the client is not aborted and continues to work.
1666 </p><div class="note"><h3 class="title">Note</h3><p>
1667 Without specifying the <code class="option">--max-tries</code> option, a transaction will
1668 never be retried after a serialization or deadlock error because its default
1669 value is 1. Use an unlimited number of tries (<code class="literal">--max-tries=0</code>)
1670 and the <code class="option">--latency-limit</code> option to limit only the maximum time
1671 of tries. You can also use the <code class="option">--time</code> option to limit the
1672 benchmark duration under an unlimited number of tries.
1674 Be careful when repeating scripts that contain multiple transactions: the
1675 script is always retried completely, so successful transactions can be
1676 performed several times.
1678 Be careful when repeating transactions with shell commands. Unlike the
1679 results of SQL commands, the results of shell commands are not rolled back,
1680 except for the variable value of the <code class="command">\setshell</code> command.
1682 The latency of a successful transaction includes the entire time of
1683 transaction execution with rollbacks and retries. The latency is measured
1684 only for successful transactions and commands but not for failed transactions
1687 The main report contains the number of failed transactions. If the
1688 <code class="option">--max-tries</code> option is not equal to 1, the main report also
1689 contains statistics related to retries: the total number of retried
1690 transactions and total number of retries. The per-script report inherits all
1691 these fields from the main report. The per-statement report displays retry
1692 statistics only if the <code class="option">--max-tries</code> option is not equal to 1.
1694 If you want to group failures by basic types in per-transaction and
1695 aggregation logs, as well as in the main and per-script reports, use the
1696 <code class="option">--failures-detailed</code> option. If you also want to distinguish
1697 all errors and failures (errors without retrying) by type including which
1698 limit for retries was exceeded and how much it was exceeded by for the
1699 serialization/deadlock failures, use the <code class="option">--verbose-errors</code>
1701 </p></div><div class="refsect2" id="id-1.9.4.11.9.10"><h3>Table Access Methods</h3><p>
1702 You may specify the <a class="link" href="tableam.html" title="Chapter 62. Table Access Method Interface Definition">Table Access Method</a>
1703 for the pgbench tables. The environment variable <code class="envar">PGOPTIONS</code>
1704 specifies database configuration options that are passed to PostgreSQL via
1705 the command line (See <a class="xref" href="config-setting.html#CONFIG-SETTING-SHELL" title="19.1.4. Parameter Interaction via the Shell">Section 19.1.4</a>).
1706 For example, a hypothetical default Table Access Method for the tables that
1707 pgbench creates called <code class="literal">wuzza</code> can be specified with:
1708 </p><pre class="programlisting">
1709 PGOPTIONS='-c default_table_access_method=wuzza'
1711 </p></div><div class="refsect2" id="id-1.9.4.11.9.11"><h3>Good Practices</h3><p>
1712 It is very easy to use <span class="application">pgbench</span> to produce completely
1713 meaningless numbers. Here are some guidelines to help you get useful
1716 In the first place, <span class="emphasis"><em>never</em></span> believe any test that runs
1717 for only a few seconds. Use the <code class="option">-t</code> or <code class="option">-T</code> option
1718 to make the run last at least a few minutes, so as to average out noise.
1719 In some cases you could need hours to get numbers that are reproducible.
1720 It's a good idea to try the test run a few times, to find out if your
1721 numbers are reproducible or not.
1723 For the default TPC-B-like test scenario, the initialization scale factor
1724 (<code class="option">-s</code>) should be at least as large as the largest number of
1725 clients you intend to test (<code class="option">-c</code>); else you'll mostly be
1726 measuring update contention. There are only <code class="option">-s</code> rows in
1727 the <code class="structname">pgbench_branches</code> table, and every transaction wants to
1728 update one of them, so <code class="option">-c</code> values in excess of <code class="option">-s</code>
1729 will undoubtedly result in lots of transactions blocked waiting for
1732 The default test scenario is also quite sensitive to how long it's been
1733 since the tables were initialized: accumulation of dead rows and dead space
1734 in the tables changes the results. To understand the results you must keep
1735 track of the total number of updates and when vacuuming happens. If
1736 autovacuum is enabled it can result in unpredictable changes in measured
1739 A limitation of <span class="application">pgbench</span> is that it can itself become
1740 the bottleneck when trying to test a large number of client sessions.
1741 This can be alleviated by running <span class="application">pgbench</span> on a different
1742 machine from the database server, although low network latency will be
1743 essential. It might even be useful to run several <span class="application">pgbench</span>
1744 instances concurrently, on several client machines, against the same
1746 </p></div><div class="refsect2" id="id-1.9.4.11.9.12"><h3>Security</h3><p>
1747 If untrusted users have access to a database that has not adopted a
1748 <a class="link" href="ddl-schemas.html#DDL-SCHEMAS-PATTERNS" title="5.10.6. Usage Patterns">secure schema usage pattern</a>,
1749 do not run <span class="application">pgbench</span> in that
1750 database. <span class="application">pgbench</span> uses unqualified names and
1751 does not manipulate the search path.
1752 </p></div></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="app-pgbasebackup.html" title="pg_basebackup">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="reference-client.html" title="PostgreSQL Client Applications">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="app-pgcombinebackup.html" title="pg_combinebackup">Next</a></td></tr><tr><td width="40%" align="left" valign="top"><span class="application">pg_basebackup</span> </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"> <span class="application">pg_combinebackup</span></td></tr></table></div></body></html>
begriffs.com / projects / ai-pg / blob