1POSTGRES(1) PostgreSQL 9.2.24 Documentation POSTGRES(1)
2
3
4
6 postgres - PostgreSQL database server
7
9 postgres [option...]
10
12 postgres is the PostgreSQL database server. In order for a client
13 application to access a database it connects (over a network or
14 locally) to a running postgres instance. The postgres instance then
15 starts a separate server process to handle the connection.
16
17 One postgres instance always manages the data of exactly one database
18 cluster. A database cluster is a collection of databases that is stored
19 at a common file system location (the “data area”). More than one
20 postgres instance can run on a system at one time, so long as they use
21 different data areas and different communication ports (see below).
22 When postgres starts it needs to know the location of the data area.
23 The location must be specified by the -D option or the PGDATA
24 environment variable; there is no default. Typically, -D or PGDATA
25 points directly to the data area directory created by initdb(1). Other
26 possible file layouts are discussed in Section 18.2, “File Locations”,
27 in the documentation.
28
29 By default postgres starts in the foreground and prints log messages to
30 the standard error stream. In practical applications postgres should be
31 started as a background process, perhaps at boot time.
32
33 The postgres command can also be called in single-user mode. The
34 primary use for this mode is during bootstrapping by initdb(1).
35 Sometimes it is used for debugging or disaster recovery; note that
36 running a single-user server is not truly suitable for debugging the
37 server, since no realistic interprocess communication and locking will
38 happen. When invoked in single-user mode from the shell, the user can
39 enter queries and the results will be printed to the screen, but in a
40 form that is more useful for developers than end users. In the
41 single-user mode, the session user will be set to the user with ID 1,
42 and implicit superuser powers are granted to this user. This user does
43 not actually have to exist, so the single-user mode can be used to
44 manually recover from certain kinds of accidental damage to the system
45 catalogs.
46
48 postgres accepts the following command-line arguments. For a detailed
49 discussion of the options consult Chapter 18, Server Configuration, in
50 the documentation. You can save typing most of these options by setting
51 up a configuration file. Some (safe) options can also be set from the
52 connecting client in an application-dependent way to apply only for
53 that session. For example, if the environment variable PGOPTIONS is
54 set, then libpq-based clients will pass that string to the server,
55 which will interpret it as postgres command-line options.
56
57 General Purpose
58 -A 0|1
59 Enables run-time assertion checks, which is a debugging aid to
60 detect programming mistakes. This option is only available if
61 assertions were enabled when PostgreSQL was compiled. If so, the
62 default is on.
63
64 -B nbuffers
65 Sets the number of shared buffers for use by the server processes.
66 The default value of this parameter is chosen automatically by
67 initdb. Specifying this option is equivalent to setting the
68 shared_buffers configuration parameter.
69
70 -c name=value
71 Sets a named run-time parameter. The configuration parameters
72 supported by PostgreSQL are described in Chapter 18, Server
73 Configuration, in the documentation. Most of the other command line
74 options are in fact short forms of such a parameter assignment. -c
75 can appear multiple times to set multiple parameters.
76
77 -C name
78 Prints the value of the named run-time parameter, and exits. (See
79 the -c option above for details.) This can be used on a running
80 server, and returns values from postgresql.conf, modified by any
81 parameters supplied in this invocation. It does not reflect
82 parameters supplied when the cluster was started.
83
84 This option is meant for other programs that interact with a server
85 instance, such as pg_ctl(1), to query configuration parameter
86 values. User-facing applications should instead use SHOW(7) or the
87 pg_settings view.
88
89 -d debug-level
90 Sets the debug level. The higher this value is set, the more
91 debugging output is written to the server log. Values are from 1 to
92 5. It is also possible to pass -d 0 for a specific session, which
93 will prevent the server log level of the parent postgres process
94 from being propagated to this session.
95
96 -D datadir
97 Specifies the file system location of the data directory or
98 configuration file(s). See Section 18.2, “File Locations”, in the
99 documentation for details.
100
101 -e
102 Sets the default date style to “European”, that is DMY ordering of
103 input date fields. This also causes the day to be printed before
104 the month in certain date output formats. See Section 8.5,
105 “Date/Time Types”, in the documentation for more information.
106
107 -F
108 Disables fsync calls for improved performance, at the risk of data
109 corruption in the event of a system crash. Specifying this option
110 is equivalent to disabling the fsync configuration parameter. Read
111 the detailed documentation before using this!
112
113 -h hostname
114 Specifies the IP host name or address on which postgres is to
115 listen for TCP/IP connections from client applications. The value
116 can also be a comma-separated list of addresses, or * to specify
117 listening on all available interfaces. An empty value specifies not
118 listening on any IP addresses, in which case only Unix-domain
119 sockets can be used to connect to the server. Defaults to listening
120 only on localhost. Specifying this option is equivalent to setting
121 the listen_addresses configuration parameter.
122
123 -i
124 Allows remote clients to connect via TCP/IP (Internet domain)
125 connections. Without this option, only local connections are
126 accepted. This option is equivalent to setting listen_addresses to
127 * in postgresql.conf or via -h.
128
129 This option is deprecated since it does not allow access to the
130 full functionality of listen_addresses. It's usually better to set
131 listen_addresses directly.
132
133 -k directory
134 Specifies the directory of the Unix-domain socket on which postgres
135 is to listen for connections from client applications. The value
136 can also be a comma-separated list of directories. An empty value
137 specifies not listening on any Unix-domain sockets, in which case
138 only TCP/IP sockets can be used to connect to the server. The
139 default value is normally /tmp, but that can be changed at build
140 time. Specifying this option is equivalent to setting the
141 unix_socket_directories configuration parameter.
142
143 -l
144 Enables secure connections using SSL. PostgreSQL must have been
145 compiled with support for SSL for this option to be available. For
146 more information on using SSL, refer to Section 17.9, “Secure
147 TCP/IP Connections with SSL”, in the documentation.
148
149 -N max-connections
150 Sets the maximum number of client connections that this server will
151 accept. The default value of this parameter is chosen automatically
152 by initdb. Specifying this option is equivalent to setting the
153 max_connections configuration parameter.
154
155 -o extra-options
156 The command-line-style options specified in extra-options are
157 passed to all server processes started by this postgres process. If
158 the option string contains any spaces, the entire string must be
159 quoted.
160
161 The use of this option is obsolete; all command-line options for
162 server processes can be specified directly on the postgres command
163 line.
164
165 -p port
166 Specifies the TCP/IP port or local Unix domain socket file
167 extension on which postgres is to listen for connections from
168 client applications. Defaults to the value of the PGPORT
169 environment variable, or if PGPORT is not set, then defaults to the
170 value established during compilation (normally 5432). If you
171 specify a port other than the default port, then all client
172 applications must specify the same port using either command-line
173 options or PGPORT.
174
175 -s
176 Print time information and other statistics at the end of each
177 command. This is useful for benchmarking or for use in tuning the
178 number of buffers.
179
180 -S work-mem
181 Specifies the amount of memory to be used by internal sorts and
182 hashes before resorting to temporary disk files. See the
183 description of the work_mem configuration parameter in Section
184 18.4.1, “Memory”, in the documentation.
185
186 -V, --version
187 Print the postgres version and exit.
188
189 --name=value
190 Sets a named run-time parameter; a shorter form of -c.
191
192 --describe-config
193 This option dumps out the server's internal configuration
194 variables, descriptions, and defaults in tab-delimited COPY format.
195 It is designed primarily for use by administration tools.
196
197 -?, --help
198 Show help about postgres command line arguments, and exit.
199
200 Semi-internal Options
201 The options described here are used mainly for debugging purposes, and
202 in some cases to assist with recovery of severely damaged databases.
203 There should be no reason to use them in a production database setup.
204 They are listed here only for use by PostgreSQL system developers.
205 Furthermore, these options might change or be removed in a future
206 release without notice.
207
208 -f { s | i | o | b | t | n | m | h }
209 Forbids the use of particular scan and join methods: s and i
210 disable sequential and index scans respectively, o, b and t disable
211 index-only scans, bitmap index scans, and TID scans respectively,
212 while n, m, and h disable nested-loop, merge and hash joins
213 respectively.
214
215 Neither sequential scans nor nested-loop joins can be disabled
216 completely; the -fs and -fn options simply discourage the optimizer
217 from using those plan types if it has any other alternative.
218
219 -n
220 This option is for debugging problems that cause a server process
221 to die abnormally. The ordinary strategy in this situation is to
222 notify all other server processes that they must terminate and then
223 reinitialize the shared memory and semaphores. This is because an
224 errant server process could have corrupted some shared state before
225 terminating. This option specifies that postgres will not
226 reinitialize shared data structures. A knowledgeable system
227 programmer can then use a debugger to examine shared memory and
228 semaphore state.
229
230 -O
231 Allows the structure of system tables to be modified. This is used
232 by initdb.
233
234 -P
235 Ignore system indexes when reading system tables, but still update
236 the indexes when modifying the tables. This is useful when
237 recovering from damaged system indexes.
238
239 -t pa[rser] | pl[anner] | e[xecutor]
240 Print timing statistics for each query relating to each of the
241 major system modules. This option cannot be used together with the
242 -s option.
243
244 -T
245 This option is for debugging problems that cause a server process
246 to die abnormally. The ordinary strategy in this situation is to
247 notify all other server processes that they must terminate and then
248 reinitialize the shared memory and semaphores. This is because an
249 errant server process could have corrupted some shared state before
250 terminating. This option specifies that postgres will stop all
251 other server processes by sending the signal SIGSTOP, but will not
252 cause them to terminate. This permits system programmers to collect
253 core dumps from all server processes by hand.
254
255 -v protocol
256 Specifies the version number of the frontend/backend protocol to be
257 used for a particular session. This option is for internal use
258 only.
259
260 -W seconds
261 A delay of this many seconds occurs when a new server process is
262 started, after it conducts the authentication procedure. This is
263 intended to give an opportunity to attach to the server process
264 with a debugger.
265
266 Options for Single-User Mode
267 The following options only apply to the single-user mode.
268
269 --single
270 Selects the single-user mode. This must be the first argument on
271 the command line.
272
273 database
274 Specifies the name of the database to be accessed. This must be the
275 last argument on the command line. If it is omitted it defaults to
276 the user name.
277
278 -E
279 Echo all commands.
280
281 -j
282 Disables use of newline as a statement delimiter.
283
284 -r filename
285 Send all server log output to filename. In normal multiuser mode,
286 this option is ignored, and stderr is used by all processes.
287
289 PGCLIENTENCODING
290 Default character encoding used by clients. (The clients can
291 override this individually.) This value can also be set in the
292 configuration file.
293
294 PGDATA
295 Default data directory location
296
297 PGDATESTYLE
298 Default value of the DateStyle run-time parameter. (The use of this
299 environment variable is deprecated.)
300
301 PGPORT
302 Default port number (preferably set in the configuration file)
303
305 A failure message mentioning semget or shmget probably indicates you
306 need to configure your kernel to provide adequate shared memory and
307 semaphores. For more discussion see Section 17.4, “Managing Kernel
308 Resources”, in the documentation. You might be able to postpone
309 reconfiguring your kernel by decreasing shared_buffers to reduce the
310 shared memory consumption of PostgreSQL, and/or by reducing
311 max_connections to reduce the semaphore consumption.
312
313 A failure message suggesting that another server is already running
314 should be checked carefully, for example by using the command
315
316 $ ps ax | grep postgres
317
318 or
319
320 $ ps -ef | grep postgres
321
322 depending on your system. If you are certain that no conflicting server
323 is running, you can remove the lock file mentioned in the message and
324 try again.
325
326 A failure message indicating inability to bind to a port might indicate
327 that that port is already in use by some non-PostgreSQL process. You
328 might also get this error if you terminate postgres and immediately
329 restart it using the same port; in this case, you must simply wait a
330 few seconds until the operating system closes the port before trying
331 again. Finally, you might get this error if you specify a port number
332 that your operating system considers to be reserved. For example, many
333 versions of Unix consider port numbers under 1024 to be “trusted” and
334 only permit the Unix superuser to access them.
335
337 The utility command pg_ctl(1) can be used to start and shut down the
338 postgres server safely and comfortably.
339
340 If at all possible, do not use SIGKILL to kill the main postgres
341 server. Doing so will prevent postgres from freeing the system
342 resources (e.g., shared memory and semaphores) that it holds before
343 terminating. This might cause problems for starting a fresh postgres
344 run.
345
346 To terminate the postgres server normally, the signals SIGTERM, SIGINT,
347 or SIGQUIT can be used. The first will wait for all clients to
348 terminate before quitting, the second will forcefully disconnect all
349 clients, and the third will quit immediately without proper shutdown,
350 resulting in a recovery run during restart.
351
352 The SIGHUP signal will reload the server configuration files. It is
353 also possible to send SIGHUP to an individual server process, but that
354 is usually not sensible.
355
356 To cancel a running query, send the SIGINT signal to the process
357 running that command.
358
359 The postgres server uses SIGTERM to tell subordinate server processes
360 to quit normally and SIGQUIT to terminate without the normal cleanup.
361 These signals should not be used by users. It is also unwise to send
362 SIGKILL to a server process — the main postgres process will interpret
363 this as a crash and will force all the sibling processes to quit as
364 part of its standard crash-recovery procedure.
365
367 The -- options will not work on FreeBSD or OpenBSD. Use -c instead.
368 This is a bug in the affected operating systems; a future release of
369 PostgreSQL will provide a workaround if this is not fixed.
370
372 To start a single-user mode server, use a command like
373
374 postgres --single -D /usr/local/pgsql/data other-options my_database
375
376 Provide the correct path to the database directory with -D, or make
377 sure that the environment variable PGDATA is set. Also specify the name
378 of the particular database you want to work in.
379
380 Normally, the single-user mode server treats newline as the command
381 entry terminator; there is no intelligence about semicolons, as there
382 is in psql. To continue a command across multiple lines, you must type
383 backslash just before each newline except the last one.
384
385 But if you use the -j command line switch, then newline does not
386 terminate command entry. In this case, the server will read the
387 standard input until the end-of-file (EOF) marker, then process the
388 input as a single command string. Backslash-newline is not treated
389 specially in this case.
390
391 To quit the session, type EOF (Control+D, usually). If you've used -j,
392 two consecutive EOFs are needed to exit.
393
394 Note that the single-user mode server does not provide sophisticated
395 line-editing features (no command history, for example). Single-User
396 mode also does not do any background processing, like automatic
397 checkpoints.
398
400 To start postgres in the background using default values, type:
401
402 $ nohup postgres >logfile 2>&1 </dev/null &
403
404 To start postgres with a specific port, e.g. 1234:
405
406 $ postgres -p 1234
407
408 To connect to this server using psql, specify this port with the -p
409 option:
410
411 $ psql -p 1234
412
413 or set the environment variable PGPORT:
414
415 $ export PGPORT=1234
416 $ psql
417
418 Named run-time parameters can be set in either of these styles:
419
420 $ postgres -c work_mem=1234
421 $ postgres --work-mem=1234
422
423 Either form overrides whatever setting might exist for work_mem in
424 postgresql.conf. Notice that underscores in parameter names can be
425 written as either underscore or dash on the command line. Except for
426 short-term experiments, it's probably better practice to edit the
427 setting in postgresql.conf than to rely on a command-line switch to set
428 a parameter.
429
431 initdb(1), pg_ctl(1)
432
433
434
435PostgreSQL 9.2.24 2017-11-06 POSTGRES(1)