1PG_RESTORE(1) PostgreSQL 12.2 Documentation PG_RESTORE(1)
2
6 pg_restore - restore a PostgreSQL database from an archive file created
7 by pg_dump
8
10 pg_restore [connection-option...] [option...] [filename]
11
13 pg_restore is a utility for restoring a PostgreSQL database from an
14 archive created by pg_dump(1) in one of the non-plain-text formats. It
15 will issue the commands necessary to reconstruct the database to the
16 state it was in at the time it was saved. The archive files also allow
17 pg_restore to be selective about what is restored, or even to reorder
18 the items prior to being restored. The archive files are designed to be
19 portable across architectures.
20 pg_restore can operate in two modes. If a database name is specified,
22 pg_restore connects to that database and restores archive contents
23 directly into the database. Otherwise, a script containing the SQL
24 commands necessary to rebuild the database is created and written to a
25 file or standard output. This script output is equivalent to the plain
26 text output format of pg_dump. Some of the options controlling the
27 output are therefore analogous to pg_dump options.
28 Obviously, pg_restore cannot restore information that is not present in
30 the archive file. For instance, if the archive was made using the “dump
31 data as INSERT commands” option, pg_restore will not be able to load
32 the data using COPY statements.
33
35 pg_restore accepts the following command line arguments.
36 filename
38 Specifies the location of the archive file (or directory, for a
39 directory-format archive) to be restored. If not specified, the
40 standard input is used.
41 -a
43 --data-only
44 Restore only the data, not the schema (data definitions). Table
45 data, large objects, and sequence values are restored, if present
46 in the archive.
47 This option is similar to, but for historical reasons not identical
49 to, specifying --section=data.
50 -c
52 --clean
53 Clean (drop) database objects before recreating them. (Unless
54 --if-exists is used, this might generate some harmless error
55 messages, if any objects were not present in the destination
56 database.)
57 -C
59 --create
60 Create the database before restoring into it. If --clean is also
61 specified, drop and recreate the target database before connecting
62 to it.
63 With --create, pg_restore also restores the database's comment if
65 any, and any configuration variable settings that are specific to
66 this database, that is, any ALTER DATABASE ... SET ... and ALTER
67 ROLE ... IN DATABASE ... SET ... commands that mention this
68 database. Access privileges for the database itself are also
69 restored, unless --no-acl is specified.
70 When this option is used, the database named with -d is used only
72 to issue the initial DROP DATABASE and CREATE DATABASE commands.
73 All data is restored into the database name that appears in the
74 archive.
75 -d dbname
77 --dbname=dbname
78 Connect to database dbname and restore directly into the database.
79 -e
81 --exit-on-error
82 Exit if an error is encountered while sending SQL commands to the
83 database. The default is to continue and to display a count of
84 errors at the end of the restoration.
85 -f filename
87 --file=filename
88 Specify output file for generated script, or for the listing when
89 used with -l. Use - for stdout.
90 -F format
92 --format=format
93 Specify format of the archive. It is not necessary to specify the
94 format, since pg_restore will determine the format automatically.
95 If specified, it can be one of the following:
96 c
98 custom
99 The archive is in the custom format of pg_dump.
100 d
102 directory
103 The archive is a directory archive.
104 t
106 tar
107 The archive is a tar archive.
108 -I index
110 --index=index
111 Restore definition of named index only. Multiple indexes may be
112 specified with multiple -I switches.
113 -j number-of-jobs
115 --jobs=number-of-jobs
116 Run the most time-consuming parts of pg_restore — those which load
117 data, create indexes, or create constraints — using multiple
118 concurrent jobs. This option can dramatically reduce the time to
119 restore a large database to a server running on a multiprocessor
120 machine.
121 Each job is one process or one thread, depending on the operating
123 system, and uses a separate connection to the server.
124 The optimal value for this option depends on the hardware setup of
126 the server, of the client, and of the network. Factors include the
127 number of CPU cores and the disk setup. A good place to start is
128 the number of CPU cores on the server, but values larger than that
129 can also lead to faster restore times in many cases. Of course,
130 values that are too high will lead to decreased performance because
131 of thrashing.
132 Only the custom and directory archive formats are supported with
134 this option. The input must be a regular file or directory (not,
135 for example, a pipe). This option is ignored when emitting a script
136 rather than connecting directly to a database server. Also,
137 multiple jobs cannot be used together with the option
138 --single-transaction.
139 -l
141 --list
142 List the table of contents of the archive. The output of this
143 operation can be used as input to the -L option. Note that if
144 filtering switches such as -n or -t are used with -l, they will
145 restrict the items listed.
146 -L list-file
148 --use-list=list-file
149 Restore only those archive elements that are listed in list-file,
150 and restore them in the order they appear in the file. Note that if
151 filtering switches such as -n or -t are used with -L, they will
152 further restrict the items restored.
153 list-file is normally created by editing the output of a previous
155 -l operation. Lines can be moved or removed, and can also be
156 commented out by placing a semicolon (;) at the start of the line.
157 See below for examples.
158 -n schema
160 --schema=schema
161 Restore only objects that are in the named schema. Multiple schemas
162 may be specified with multiple -n switches. This can be combined
163 with the -t option to restore just a specific table.
164 -N schema
166 --exclude-schema=schema
167 Do not restore objects that are in the named schema. Multiple
168 schemas to be excluded may be specified with multiple -N switches.
169 When both -n and -N are given for the same schema name, the -N
171 switch wins and the schema is excluded.
172 -O
174 --no-owner
175 Do not output commands to set ownership of objects to match the
176 original database. By default, pg_restore issues ALTER OWNER or SET
177 SESSION AUTHORIZATION statements to set ownership of created schema
178 elements. These statements will fail unless the initial connection
179 to the database is made by a superuser (or the same user that owns
180 all of the objects in the script). With -O, any user name can be
181 used for the initial connection, and this user will own all the
182 created objects.
183 -P function-name(argtype [, ...])
185 --function=function-name(argtype [, ...])
186 Restore the named function only. Be careful to spell the function
187 name and arguments exactly as they appear in the dump file's table
188 of contents. Multiple functions may be specified with multiple -P
189 switches.
190 -R
192 --no-reconnect
193 This option is obsolete but still accepted for backwards
194 compatibility.
195 -s
197 --schema-only
198 Restore only the schema (data definitions), not data, to the extent
199 that schema entries are present in the archive.
200 This option is the inverse of --data-only. It is similar to, but
202 for historical reasons not identical to, specifying
203 --section=pre-data --section=post-data.
204 (Do not confuse this with the --schema option, which uses the word
206 “schema” in a different meaning.)
207 -S username
209 --superuser=username
210 Specify the superuser user name to use when disabling triggers.
211 This is relevant only if --disable-triggers is used.
212 -t table
214 --table=table
215 Restore definition and/or data of only the named table. For this
216 purpose, “table” includes views, materialized views, sequences, and
217 foreign tables. Multiple tables can be selected by writing multiple
218 -t switches. This option can be combined with the -n option to
219 specify table(s) in a particular schema.
220 Note
222 When -t is specified, pg_restore makes no attempt to restore
223 any other database objects that the selected table(s) might
224 depend upon. Therefore, there is no guarantee that a
225 specific-table restore into a clean database will succeed.
226 Note
228 This flag does not behave identically to the -t flag of
229 pg_dump. There is not currently any provision for wild-card
230 matching in pg_restore, nor can you include a schema name
231 within its -t. And, while pg_dump's -t flag will also dump
232 subsidiary objects (such as indexes) of the selected table(s),
233 pg_restore's -t flag does not include such subsidiary objects.
234 Note
236 In versions prior to PostgreSQL 9.6, this flag matched only
237 tables, not any other type of relation.
238 -T trigger
240 --trigger=trigger
241 Restore named trigger only. Multiple triggers may be specified with
242 multiple -T switches.
243 -v
245 --verbose
246 Specifies verbose mode.
247 -V
249 --version
250 Print the pg_restore version and exit.
251 -x
253 --no-privileges
254 --no-acl
255 Prevent restoration of access privileges (grant/revoke commands).
256 -1
258 --single-transaction
259 Execute the restore as a single transaction (that is, wrap the
260 emitted commands in BEGIN/COMMIT). This ensures that either all the
261 commands complete successfully, or no changes are applied. This
262 option implies --exit-on-error.
263 --disable-triggers
265 This option is relevant only when performing a data-only restore.
266 It instructs pg_restore to execute commands to temporarily disable
267 triggers on the target tables while the data is reloaded. Use this
268 if you have referential integrity checks or other triggers on the
269 tables that you do not want to invoke during data reload.
270 Presently, the commands emitted for --disable-triggers must be done
272 as superuser. So you should also specify a superuser name with -S
273 or, preferably, run pg_restore as a PostgreSQL superuser.
274 --enable-row-security
276 This option is relevant only when restoring the contents of a table
277 which has row security. By default, pg_restore will set
278 row_security to off, to ensure that all data is restored in to the
279 table. If the user does not have sufficient privileges to bypass
280 row security, then an error is thrown. This parameter instructs
281 pg_restore to set row_security to on instead, allowing the user to
282 attempt to restore the contents of the table with row security
283 enabled. This might still fail if the user does not have the right
284 to insert the rows from the dump into the table.
285 Note that this option currently also requires the dump be in INSERT
287 format, as COPY FROM does not support row security.
288 --if-exists
290 Use conditional commands (i.e. add an IF EXISTS clause) to drop
291 database objects. This option is not valid unless --clean is also
292 specified.
293 --no-comments
295 Do not output commands to restore comments, even if the archive
296 contains them.
297 --no-data-for-failed-tables
299 By default, table data is restored even if the creation command for
300 the table failed (e.g., because it already exists). With this
301 option, data for such a table is skipped. This behavior is useful
302 if the target database already contains the desired table contents.
303 For example, auxiliary tables for PostgreSQL extensions such as
304 PostGIS might already be loaded in the target database; specifying
305 this option prevents duplicate or obsolete data from being loaded
306 into them.
307 This option is effective only when restoring directly into a
309 database, not when producing SQL script output.
310 --no-publications
312 Do not output commands to restore publications, even if the archive
313 contains them.
314 --no-security-labels
316 Do not output commands to restore security labels, even if the
317 archive contains them.
318 --no-subscriptions
320 Do not output commands to restore subscriptions, even if the
321 archive contains them.
322 --no-tablespaces
324 Do not output commands to select tablespaces. With this option, all
325 objects will be created in whichever tablespace is the default
326 during restore.
327 --section=sectionname
329 Only restore the named section. The section name can be pre-data,
330 data, or post-data. This option can be specified more than once to
331 select multiple sections. The default is to restore all sections.
332 The data section contains actual table data as well as large-object
334 definitions. Post-data items consist of definitions of indexes,
335 triggers, rules and constraints other than validated check
336 constraints. Pre-data items consist of all other data definition
337 items.
338 --strict-names
340 Require that each schema (-n/--schema) and table (-t/--table)
341 qualifier match at least one schema/table in the backup file.
342 --use-set-session-authorization
344 Output SQL-standard SET SESSION AUTHORIZATION commands instead of
345 ALTER OWNER commands to determine object ownership. This makes the
346 dump more standards-compatible, but depending on the history of the
347 objects in the dump, might not restore properly.
348 -?
350 --help
351 Show help about pg_restore command line arguments, and exit.
352 pg_restore also accepts the following command line arguments for
354 connection parameters:
355 -h host
357 --host=host
358 Specifies the host name of the machine on which the server is
359 running. If the value begins with a slash, it is used as the
360 directory for the Unix domain socket. The default is taken from the
361 PGHOST environment variable, if set, else a Unix domain socket
362 connection is attempted.
363 -p port
365 --port=port
366 Specifies the TCP port or local Unix domain socket file extension
367 on which the server is listening for connections. Defaults to the
368 PGPORT environment variable, if set, or a compiled-in default.
369 -U username
371 --username=username
372 User name to connect as.
373 -w
375 --no-password
376 Never issue a password prompt. If the server requires password
377 authentication and a password is not available by other means such
378 as a .pgpass file, the connection attempt will fail. This option
379 can be useful in batch jobs and scripts where no user is present to
380 enter a password.
381 -W
383 --password
384 Force pg_restore to prompt for a password before connecting to a
385 database.
386 This option is never essential, since pg_restore will automatically
388 prompt for a password if the server demands password
389 authentication. However, pg_restore will waste a connection attempt
390 finding out that the server wants a password. In some cases it is
391 worth typing -W to avoid the extra connection attempt.
392 --role=rolename
394 Specifies a role name to be used to perform the restore. This
395 option causes pg_restore to issue a SET ROLE rolename command after
396 connecting to the database. It is useful when the authenticated
397 user (specified by -U) lacks privileges needed by pg_restore, but
398 can switch to a role with the required rights. Some installations
399 have a policy against logging in directly as a superuser, and use
400 of this option allows restores to be performed without violating
401 the policy.
402
404 PGHOST
405 PGOPTIONS
406 PGPORT
407 PGUSER
408 Default connection parameters
409 PG_COLOR
411 Specifies whether to use color in diagnostics messages. Possible
412 values are always, auto, never.
413 This utility, like most other PostgreSQL utilities, also uses the
415 environment variables supported by libpq (see Section 33.14). However,
416 it does not read PGDATABASE when a database name is not supplied.
417
419 When a direct database connection is specified using the -d option,
420 pg_restore internally executes SQL statements. If you have problems
421 running pg_restore, make sure you are able to select information from
422 the database using, for example, psql(1). Also, any default connection
423 settings and environment variables used by the libpq front-end library
424 will apply.
425
427 If your installation has any local additions to the template1 database,
428 be careful to load the output of pg_restore into a truly empty
429 database; otherwise you are likely to get errors due to duplicate
430 definitions of the added objects. To make an empty database without any
431 local additions, copy from template0 not template1, for example:
432 CREATE DATABASE foo WITH TEMPLATE template0;
434 The limitations of pg_restore are detailed below.
436 · When restoring data to a pre-existing table and the option
438 --disable-triggers is used, pg_restore emits commands to disable
439 triggers on user tables before inserting the data, then emits
440 commands to re-enable them after the data has been inserted. If the
441 restore is stopped in the middle, the system catalogs might be left
442 in the wrong state.
443 · pg_restore cannot restore large objects selectively; for instance,
445 only those for a specific table. If an archive contains large
446 objects, then all large objects will be restored, or none of them
447 if they are excluded via -L, -t, or other options.
448 See also the pg_dump(1) documentation for details on limitations of
450 pg_dump.
451 Once restored, it is wise to run ANALYZE on each restored table so the
453 optimizer has useful statistics; see Section 24.1.3 and Section 24.1.6
454 for more information.
455
457 Assume we have dumped a database called mydb into a custom-format dump
458 file:
459 $ pg_dump -Fc mydb > db.dump
461 To drop the database and recreate it from the dump:
463 $ dropdb mydb
465 $ pg_restore -C -d postgres db.dump
466 The database named in the -d switch can be any database existing in the
468 cluster; pg_restore only uses it to issue the CREATE DATABASE command
469 for mydb. With -C, data is always restored into the database name that
470 appears in the dump file.
471 To reload the dump into a new database called newdb:
473 $ createdb -T template0 newdb
475 $ pg_restore -d newdb db.dump
476 Notice we don't use -C, and instead connect directly to the database to
478 be restored into. Also note that we clone the new database from
479 template0 not template1, to ensure it is initially empty.
480 To reorder database items, it is first necessary to dump the table of
482 contents of the archive:
483 $ pg_restore -l db.dump > db.list
485 The listing file consists of a header and one line for each item, e.g.:
487 ;
489 ; Archive created at Mon Sep 14 13:55:39 2009
490 ; dbname: DBDEMOS
491 ; TOC Entries: 81
492 ; Compression: 9
493 ; Dump Version: 1.10-0
494 ; Format: CUSTOM
495 ; Integer: 4 bytes
496 ; Offset: 8 bytes
497 ; Dumped from database version: 8.3.5
498 ; Dumped by pg_dump version: 8.3.8
499 ;
500 ;
501 ; Selected TOC Entries:
502 ;
503 3; 2615 2200 SCHEMA - public pasha
504 1861; 0 0 COMMENT - SCHEMA public pasha
505 1862; 0 0 ACL - public pasha
506 317; 1247 17715 TYPE public composite pasha
507 319; 1247 25899 DOMAIN public domain0 pasha
508 Semicolons start a comment, and the numbers at the start of lines refer
510 to the internal archive ID assigned to each item.
511 Lines in the file can be commented out, deleted, and reordered. For
513 example:
514 10; 145433 TABLE map_resolutions postgres
516 ;2; 145344 TABLE species postgres
517 ;4; 145359 TABLE nt_header postgres
518 6; 145402 TABLE species_records postgres
519 ;8; 145416 TABLE ss_old postgres
520 could be used as input to pg_restore and would only restore items 10
522 and 6, in that order:
523 $ pg_restore -L db.list db.dump
525
527 pg_dump(1), pg_dumpall(1), psql(1)
528PostgreSQL 12.2 2020 PG_RESTORE(1)