1PG_DUMP(1) PostgreSQL 13.4 Documentation PG_DUMP(1)
2
6 pg_dump - extract a PostgreSQL database into a script file or other
7 archive file
8
10 pg_dump [connection-option...] [option...] [dbname]
11
13 pg_dump is a utility for backing up a PostgreSQL database. It makes
14 consistent backups even if the database is being used concurrently.
15 pg_dump does not block other users accessing the database (readers or
16 writers).
17 pg_dump only dumps a single database. To back up an entire cluster, or
19 to back up global objects that are common to all databases in a cluster
20 (such as roles and tablespaces), use pg_dumpall(1).
21 Dumps can be output in script or archive file formats. Script dumps are
23 plain-text files containing the SQL commands required to reconstruct
24 the database to the state it was in at the time it was saved. To
25 restore from such a script, feed it to psql(1). Script files can be
26 used to reconstruct the database even on other machines and other
27 architectures; with some modifications, even on other SQL database
28 products.
29 The alternative archive file formats must be used with pg_restore(1) to
31 rebuild the database. They allow pg_restore to be selective about what
32 is restored, or even to reorder the items prior to being restored. The
33 archive file formats are designed to be portable across architectures.
34 When used with one of the archive file formats and combined with
36 pg_restore, pg_dump provides a flexible archival and transfer
37 mechanism. pg_dump can be used to backup an entire database, then
38 pg_restore can be used to examine the archive and/or select which parts
39 of the database are to be restored. The most flexible output file
40 formats are the “custom” format (-Fc) and the “directory” format (-Fd).
41 They allow for selection and reordering of all archived items, support
42 parallel restoration, and are compressed by default. The “directory”
43 format is the only format that supports parallel dumps.
44 While running pg_dump, one should examine the output for any warnings
46 (printed on standard error), especially in light of the limitations
47 listed below.
48
50 The following command-line options control the content and format of
51 the output.
52 dbname
54 Specifies the name of the database to be dumped. If this is not
55 specified, the environment variable PGDATABASE is used. If that is
56 not set, the user name specified for the connection is used.
57 -a
59 --data-only
60 Dump only the data, not the schema (data definitions). Table data,
61 large objects, and sequence values are dumped.
62 This option is similar to, but for historical reasons not identical
64 to, specifying --section=data.
65 -b
67 --blobs
68 Include large objects in the dump. This is the default behavior
69 except when --schema, --table, or --schema-only is specified. The
70 -b switch is therefore only useful to add large objects to dumps
71 where a specific schema or table has been requested. Note that
72 blobs are considered data and therefore will be included when
73 --data-only is used, but not when --schema-only is.
74 -B
76 --no-blobs
77 Exclude large objects in the dump.
78 When both -b and -B are given, the behavior is to output large
80 objects, when data is being dumped, see the -b documentation.
81 -c
83 --clean
84 Output commands to clean (drop) database objects prior to
85 outputting the commands for creating them. (Unless --if-exists is
86 also specified, restore might generate some harmless error
87 messages, if any objects were not present in the destination
88 database.)
89 This option is ignored when emitting an archive (non-text) output
91 file. For the archive formats, you can specify the option when you
92 call pg_restore.
93 -C
95 --create
96 Begin the output with a command to create the database itself and
97 reconnect to the created database. (With a script of this form, it
98 doesn't matter which database in the destination installation you
99 connect to before running the script.) If --clean is also
100 specified, the script drops and recreates the target database
101 before reconnecting to it.
102 With --create, the output also includes the database's comment if
104 any, and any configuration variable settings that are specific to
105 this database, that is, any ALTER DATABASE ... SET ... and ALTER
106 ROLE ... IN DATABASE ... SET ... commands that mention this
107 database. Access privileges for the database itself are also
108 dumped, unless --no-acl is specified.
109 This option is ignored when emitting an archive (non-text) output
111 file. For the archive formats, you can specify the option when you
112 call pg_restore.
113 -E encoding
115 --encoding=encoding
116 Create the dump in the specified character set encoding. By
117 default, the dump is created in the database encoding. (Another way
118 to get the same result is to set the PGCLIENTENCODING environment
119 variable to the desired dump encoding.)
120 -f file
122 --file=file
123 Send output to the specified file. This parameter can be omitted
124 for file based output formats, in which case the standard output is
125 used. It must be given for the directory output format however,
126 where it specifies the target directory instead of a file. In this
127 case the directory is created by pg_dump and must not exist before.
128 -F format
130 --format=format
131 Selects the format of the output. format can be one of the
132 following:
133 p
135 plain
136 Output a plain-text SQL script file (the default).
137 c
139 custom
140 Output a custom-format archive suitable for input into
141 pg_restore. Together with the directory output format, this is
142 the most flexible output format in that it allows manual
143 selection and reordering of archived items during restore. This
144 format is also compressed by default.
145 d
147 directory
148 Output a directory-format archive suitable for input into
149 pg_restore. This will create a directory with one file for each
150 table and blob being dumped, plus a so-called Table of Contents
151 file describing the dumped objects in a machine-readable format
152 that pg_restore can read. A directory format archive can be
153 manipulated with standard Unix tools; for example, files in an
154 uncompressed archive can be compressed with the gzip tool. This
155 format is compressed by default and also supports parallel
156 dumps.
157 t
159 tar
160 Output a tar-format archive suitable for input into pg_restore.
161 The tar format is compatible with the directory format:
162 extracting a tar-format archive produces a valid
163 directory-format archive. However, the tar format does not
164 support compression. Also, when using tar format the relative
165 order of table data items cannot be changed during restore.
166 -j njobs
168 --jobs=njobs
169 Run the dump in parallel by dumping njobs tables simultaneously.
170 This option may reduce the time needed to perform the dump but it
171 also increases the load on the database server. You can only use
172 this option with the directory output format because this is the
173 only output format where multiple processes can write their data at
174 the same time.
175 pg_dump will open njobs + 1 connections to the database, so make
177 sure your max_connections setting is high enough to accommodate all
178 connections.
179 Requesting exclusive locks on database objects while running a
181 parallel dump could cause the dump to fail. The reason is that the
182 pg_dump master process requests shared locks on the objects that
183 the worker processes are going to dump later in order to make sure
184 that nobody deletes them and makes them go away while the dump is
185 running. If another client then requests an exclusive lock on a
186 table, that lock will not be granted but will be queued waiting for
187 the shared lock of the master process to be released. Consequently
188 any other access to the table will not be granted either and will
189 queue after the exclusive lock request. This includes the worker
190 process trying to dump the table. Without any precautions this
191 would be a classic deadlock situation. To detect this conflict, the
192 pg_dump worker process requests another shared lock using the
193 NOWAIT option. If the worker process is not granted this shared
194 lock, somebody else must have requested an exclusive lock in the
195 meantime and there is no way to continue with the dump, so pg_dump
196 has no choice but to abort the dump.
197 For a consistent backup, the database server needs to support
199 synchronized snapshots, a feature that was introduced in PostgreSQL
200 9.2 for primary servers and 10 for standbys. With this feature,
201 database clients can ensure they see the same data set even though
202 they use different connections. pg_dump -j uses multiple database
203 connections; it connects to the database once with the master
204 process and once again for each worker job. Without the
205 synchronized snapshot feature, the different worker jobs wouldn't
206 be guaranteed to see the same data in each connection, which could
207 lead to an inconsistent backup.
208 If you want to run a parallel dump of a pre-9.2 server, you need to
210 make sure that the database content doesn't change from between the
211 time the master connects to the database until the last worker job
212 has connected to the database. The easiest way to do this is to
213 halt any data modifying processes (DDL and DML) accessing the
214 database before starting the backup. You also need to specify the
215 --no-synchronized-snapshots parameter when running pg_dump -j
216 against a pre-9.2 PostgreSQL server.
217 -n pattern
219 --schema=pattern
220 Dump only schemas matching pattern; this selects both the schema
221 itself, and all its contained objects. When this option is not
222 specified, all non-system schemas in the target database will be
223 dumped. Multiple schemas can be selected by writing multiple -n
224 switches. The pattern parameter is interpreted as a pattern
225 according to the same rules used by psql's \d commands (see
226 Patterns below), so multiple schemas can also be selected by
227 writing wildcard characters in the pattern. When using wildcards,
228 be careful to quote the pattern if needed to prevent the shell from
229 expanding the wildcards; see Examples below.
230 Note
232 When -n is specified, pg_dump makes no attempt to dump any
233 other database objects that the selected schema(s) might depend
234 upon. Therefore, there is no guarantee that the results of a
235 specific-schema dump can be successfully restored by themselves
236 into a clean database.
237 Note
239 Non-schema objects such as blobs are not dumped when -n is
240 specified. You can add blobs back to the dump with the --blobs
241 switch.
242 -N pattern
244 --exclude-schema=pattern
245 Do not dump any schemas matching pattern. The pattern is
246 interpreted according to the same rules as for -n. -N can be given
247 more than once to exclude schemas matching any of several patterns.
248 When both -n and -N are given, the behavior is to dump just the
250 schemas that match at least one -n switch but no -N switches. If -N
251 appears without -n, then schemas matching -N are excluded from what
252 is otherwise a normal dump.
253 -O
255 --no-owner
256 Do not output commands to set ownership of objects to match the
257 original database. By default, pg_dump issues ALTER OWNER or SET
258 SESSION AUTHORIZATION statements to set ownership of created
259 database objects. These statements will fail when the script is run
260 unless it is started by a superuser (or the same user that owns all
261 of the objects in the script). To make a script that can be
262 restored by any user, but will give that user ownership of all the
263 objects, specify -O.
264 This option is ignored when emitting an archive (non-text) output
266 file. For the archive formats, you can specify the option when you
267 call pg_restore.
268 -R
270 --no-reconnect
271 This option is obsolete but still accepted for backwards
272 compatibility.
273 -s
275 --schema-only
276 Dump only the object definitions (schema), not data.
277 This option is the inverse of --data-only. It is similar to, but
279 for historical reasons not identical to, specifying
280 --section=pre-data --section=post-data.
281 (Do not confuse this with the --schema option, which uses the word
283 “schema” in a different meaning.)
284 To exclude table data for only a subset of tables in the database,
286 see --exclude-table-data.
287 -S username
289 --superuser=username
290 Specify the superuser user name to use when disabling triggers.
291 This is relevant only if --disable-triggers is used. (Usually, it's
292 better to leave this out, and instead start the resulting script as
293 superuser.)
294 -t pattern
296 --table=pattern
297 Dump only tables with names matching pattern. Multiple tables can
298 be selected by writing multiple -t switches. The pattern parameter
299 is interpreted as a pattern according to the same rules used by
300 psql's \d commands (see Patterns below), so multiple tables can
301 also be selected by writing wildcard characters in the pattern.
302 When using wildcards, be careful to quote the pattern if needed to
303 prevent the shell from expanding the wildcards; see Examples below.
304 As well as tables, this option can be used to dump the definition
306 of matching views, materialized views, foreign tables, and
307 sequences. It will not dump the contents of views or materialized
308 views, and the contents of foreign tables will only be dumped if
309 the corresponding foreign server is specified with
310 --include-foreign-data.
311 The -n and -N switches have no effect when -t is used, because
313 tables selected by -t will be dumped regardless of those switches,
314 and non-table objects will not be dumped.
315 Note
317 When -t is specified, pg_dump makes no attempt to dump any
318 other database objects that the selected table(s) might depend
319 upon. Therefore, there is no guarantee that the results of a
320 specific-table dump can be successfully restored by themselves
321 into a clean database.
322 Note
324 The behavior of the -t switch is not entirely upward compatible
325 with pre-8.2 PostgreSQL versions. Formerly, writing -t tab
326 would dump all tables named tab, but now it just dumps
327 whichever one is visible in your default search path. To get
328 the old behavior you can write -t '*.tab'. Also, you must write
329 something like -t sch.tab to select a table in a particular
330 schema, rather than the old locution of -n sch -t tab.
331 -T pattern
333 --exclude-table=pattern
334 Do not dump any tables matching pattern. The pattern is interpreted
335 according to the same rules as for -t. -T can be given more than
336 once to exclude tables matching any of several patterns.
337 When both -t and -T are given, the behavior is to dump just the
339 tables that match at least one -t switch but no -T switches. If -T
340 appears without -t, then tables matching -T are excluded from what
341 is otherwise a normal dump.
342 -v
344 --verbose
345 Specifies verbose mode. This will cause pg_dump to output detailed
346 object comments and start/stop times to the dump file, and progress
347 messages to standard error.
348 -V
350 --version
351 Print the pg_dump version and exit.
352 -x
354 --no-privileges
355 --no-acl
356 Prevent dumping of access privileges (grant/revoke commands).
357 -Z 0..9
359 --compress=0..9
360 Specify the compression level to use. Zero means no compression.
361 For the custom and directory archive formats, this specifies
362 compression of individual table-data segments, and the default is
363 to compress at a moderate level. For plain text output, setting a
364 nonzero compression level causes the entire output file to be
365 compressed, as though it had been fed through gzip; but the default
366 is not to compress. The tar archive format currently does not
367 support compression at all.
368 --binary-upgrade
370 This option is for use by in-place upgrade utilities. Its use for
371 other purposes is not recommended or supported. The behavior of the
372 option may change in future releases without notice.
373 --column-inserts
375 --attribute-inserts
376 Dump data as INSERT commands with explicit column names (INSERT
377 INTO table (column, ...) VALUES ...). This will make restoration
378 very slow; it is mainly useful for making dumps that can be loaded
379 into non-PostgreSQL databases. Any error during reloading will
380 cause only rows that are part of the problematic INSERT to be lost,
381 rather than the entire table contents.
382 --disable-dollar-quoting
384 This option disables the use of dollar quoting for function bodies,
385 and forces them to be quoted using SQL standard string syntax.
386 --disable-triggers
388 This option is relevant only when creating a data-only dump. It
389 instructs pg_dump to include commands to temporarily disable
390 triggers on the target tables while the data is reloaded. Use this
391 if you have referential integrity checks or other triggers on the
392 tables that you do not want to invoke during data reload.
393 Presently, the commands emitted for --disable-triggers must be done
395 as superuser. So, you should also specify a superuser name with -S,
396 or preferably be careful to start the resulting script as a
397 superuser.
398 This option is ignored when emitting an archive (non-text) output
400 file. For the archive formats, you can specify the option when you
401 call pg_restore.
402 --enable-row-security
404 This option is relevant only when dumping the contents of a table
405 which has row security. By default, pg_dump will set row_security
406 to off, to ensure that all data is dumped from the table. If the
407 user does not have sufficient privileges to bypass row security,
408 then an error is thrown. This parameter instructs pg_dump to set
409 row_security to on instead, allowing the user to dump the parts of
410 the contents of the table that they have access to.
411 Note that if you use this option currently, you probably also want
413 the dump be in INSERT format, as the COPY FROM during restore does
414 not support row security.
415 --exclude-table-data=pattern
417 Do not dump data for any tables matching pattern. The pattern is
418 interpreted according to the same rules as for -t.
419 --exclude-table-data can be given more than once to exclude tables
420 matching any of several patterns. This option is useful when you
421 need the definition of a particular table even though you do not
422 need the data in it.
423 To exclude data for all tables in the database, see --schema-only.
425 --extra-float-digits=ndigits
427 Use the specified value of extra_float_digits when dumping
428 floating-point data, instead of the maximum available precision.
429 Routine dumps made for backup purposes should not use this option.
430 --if-exists
432 Use conditional commands (i.e., add an IF EXISTS clause) when
433 cleaning database objects. This option is not valid unless --clean
434 is also specified.
435 --include-foreign-data=foreignserver
437 Dump the data for any foreign table with a foreign server matching
438 foreignserver pattern. Multiple foreign servers can be selected by
439 writing multiple --include-foreign-data switches. Also, the
440 foreignserver parameter is interpreted as a pattern according to
441 the same rules used by psql's \d commands (see Patterns below), so
442 multiple foreign servers can also be selected by writing wildcard
443 characters in the pattern. When using wildcards, be careful to
444 quote the pattern if needed to prevent the shell from expanding the
445 wildcards; see Examples below. The only exception is that an empty
446 pattern is disallowed.
447 Note
449 When --include-foreign-data is specified, pg_dump does not
450 check that the foreign table is writable. Therefore, there is
451 no guarantee that the results of a foreign table dump can be
452 successfully restored.
453 --inserts
455 Dump data as INSERT commands (rather than COPY). This will make
456 restoration very slow; it is mainly useful for making dumps that
457 can be loaded into non-PostgreSQL databases. Any error during
458 reloading will cause only rows that are part of the problematic
459 INSERT to be lost, rather than the entire table contents. Note that
460 the restore might fail altogether if you have rearranged column
461 order. The --column-inserts option is safe against column order
462 changes, though even slower.
463 --load-via-partition-root
465 When dumping data for a table partition, make the COPY or INSERT
466 statements target the root of the partitioning hierarchy that
467 contains it, rather than the partition itself. This causes the
468 appropriate partition to be re-determined for each row when the
469 data is loaded. This may be useful when reloading data on a server
470 where rows do not always fall into the same partitions as they did
471 on the original server. That could happen, for example, if the
472 partitioning column is of type text and the two systems have
473 different definitions of the collation used to sort the
474 partitioning column.
475 It is best not to use parallelism when restoring from an archive
477 made with this option, because pg_restore will not know exactly
478 which partition(s) a given archive data item will load data into.
479 This could result in inefficiency due to lock conflicts between
480 parallel jobs, or perhaps even reload failures due to foreign key
481 constraints being set up before all the relevant data is loaded.
482 --lock-wait-timeout=timeout
484 Do not wait forever to acquire shared table locks at the beginning
485 of the dump. Instead fail if unable to lock a table within the
486 specified timeout. The timeout may be specified in any of the
487 formats accepted by SET statement_timeout. (Allowed formats vary
488 depending on the server version you are dumping from, but an
489 integer number of milliseconds is accepted by all versions.)
490 --no-comments
492 Do not dump comments.
493 --no-publications
495 Do not dump publications.
496 --no-security-labels
498 Do not dump security labels.
499 --no-subscriptions
501 Do not dump subscriptions.
502 --no-sync
504 By default, pg_dump will wait for all files to be written safely to
505 disk. This option causes pg_dump to return without waiting, which
506 is faster, but means that a subsequent operating system crash can
507 leave the dump corrupt. Generally, this option is useful for
508 testing but should not be used when dumping data from production
509 installation.
510 --no-synchronized-snapshots
512 This option allows running pg_dump -j against a pre-9.2 server, see
513 the documentation of the -j parameter for more details.
514 --no-tablespaces
516 Do not output commands to select tablespaces. With this option, all
517 objects will be created in whichever tablespace is the default
518 during restore.
519 This option is ignored when emitting an archive (non-text) output
521 file. For the archive formats, you can specify the option when you
522 call pg_restore.
523 --no-unlogged-table-data
525 Do not dump the contents of unlogged tables. This option has no
526 effect on whether or not the table definitions (schema) are dumped;
527 it only suppresses dumping the table data. Data in unlogged tables
528 is always excluded when dumping from a standby server.
529 --on-conflict-do-nothing
531 Add ON CONFLICT DO NOTHING to INSERT commands. This option is not
532 valid unless --inserts, --column-inserts or --rows-per-insert is
533 also specified.
534 --quote-all-identifiers
536 Force quoting of all identifiers. This option is recommended when
537 dumping a database from a server whose PostgreSQL major version is
538 different from pg_dump's, or when the output is intended to be
539 loaded into a server of a different major version. By default,
540 pg_dump quotes only identifiers that are reserved words in its own
541 major version. This sometimes results in compatibility issues when
542 dealing with servers of other versions that may have slightly
543 different sets of reserved words. Using --quote-all-identifiers
544 prevents such issues, at the price of a harder-to-read dump script.
545 --rows-per-insert=nrows
547 Dump data as INSERT commands (rather than COPY). Controls the
548 maximum number of rows per INSERT command. The value specified must
549 be a number greater than zero. Any error during reloading will
550 cause only rows that are part of the problematic INSERT to be lost,
551 rather than the entire table contents.
552 --section=sectionname
554 Only dump the named section. The section name can be pre-data,
555 data, or post-data. This option can be specified more than once to
556 select multiple sections. The default is to dump all sections.
557 The data section contains actual table data, large-object contents,
559 and sequence values. Post-data items include definitions of
560 indexes, triggers, rules, and constraints other than validated
561 check constraints. Pre-data items include all other data definition
562 items.
563 --serializable-deferrable
565 Use a serializable transaction for the dump, to ensure that the
566 snapshot used is consistent with later database states; but do this
567 by waiting for a point in the transaction stream at which no
568 anomalies can be present, so that there isn't a risk of the dump
569 failing or causing other transactions to roll back with a
570 serialization_failure. See Chapter 13 for more information about
571 transaction isolation and concurrency control.
572 This option is not beneficial for a dump which is intended only for
574 disaster recovery. It could be useful for a dump used to load a
575 copy of the database for reporting or other read-only load sharing
576 while the original database continues to be updated. Without it the
577 dump may reflect a state which is not consistent with any serial
578 execution of the transactions eventually committed. For example, if
579 batch processing techniques are used, a batch may show as closed in
580 the dump without all of the items which are in the batch appearing.
581 This option will make no difference if there are no read-write
583 transactions active when pg_dump is started. If read-write
584 transactions are active, the start of the dump may be delayed for
585 an indeterminate length of time. Once running, performance with or
586 without the switch is the same.
587 --snapshot=snapshotname
589 Use the specified synchronized snapshot when making a dump of the
590 database (see Table 9.88 for more details).
591 This option is useful when needing to synchronize the dump with a
593 logical replication slot (see Chapter 48) or with a concurrent
594 session.
595 In the case of a parallel dump, the snapshot name defined by this
597 option is used rather than taking a new snapshot.
598 --strict-names
600 Require that each schema (-n/--schema) and table (-t/--table)
601 qualifier match at least one schema/table in the database to be
602 dumped. Note that if none of the schema/table qualifiers find
603 matches, pg_dump will generate an error even without
604 --strict-names.
605 This option has no effect on -N/--exclude-schema,
607 -T/--exclude-table, or --exclude-table-data. An exclude pattern
608 failing to match any objects is not considered an error.
609 --use-set-session-authorization
611 Output SQL-standard SET SESSION AUTHORIZATION commands instead of
612 ALTER OWNER commands to determine object ownership. This makes the
613 dump more standards-compatible, but depending on the history of the
614 objects in the dump, might not restore properly. Also, a dump using
615 SET SESSION AUTHORIZATION will certainly require superuser
616 privileges to restore correctly, whereas ALTER OWNER requires
617 lesser privileges.
618 -?
620 --help
621 Show help about pg_dump command line arguments, and exit.
622 The following command-line options control the database connection
624 parameters.
625 -d dbname
627 --dbname=dbname
628 Specifies the name of the database to connect to. This is
629 equivalent to specifying dbname as the first non-option argument on
630 the command line. The dbname can be a connection string. If so,
631 connection string parameters will override any conflicting command
632 line options.
633 -h host
635 --host=host
636 Specifies the host name of the machine on which the server is
637 running. If the value begins with a slash, it is used as the
638 directory for the Unix domain socket. The default is taken from the
639 PGHOST environment variable, if set, else a Unix domain socket
640 connection is attempted.
641 -p port
643 --port=port
644 Specifies the TCP port or local Unix domain socket file extension
645 on which the server is listening for connections. Defaults to the
646 PGPORT environment variable, if set, or a compiled-in default.
647 -U username
649 --username=username
650 User name to connect as.
651 -w
653 --no-password
654 Never issue a password prompt. If the server requires password
655 authentication and a password is not available by other means such
656 as a .pgpass file, the connection attempt will fail. This option
657 can be useful in batch jobs and scripts where no user is present to
658 enter a password.
659 -W
661 --password
662 Force pg_dump to prompt for a password before connecting to a
663 database.
664 This option is never essential, since pg_dump will automatically
666 prompt for a password if the server demands password
667 authentication. However, pg_dump will waste a connection attempt
668 finding out that the server wants a password. In some cases it is
669 worth typing -W to avoid the extra connection attempt.
670 --role=rolename
672 Specifies a role name to be used to create the dump. This option
673 causes pg_dump to issue a SET ROLE rolename command after
674 connecting to the database. It is useful when the authenticated
675 user (specified by -U) lacks privileges needed by pg_dump, but can
676 switch to a role with the required rights. Some installations have
677 a policy against logging in directly as a superuser, and use of
678 this option allows dumps to be made without violating the policy.
679
681 PGDATABASE
682 PGHOST
683 PGOPTIONS
684 PGPORT
685 PGUSER
686 Default connection parameters.
687 PG_COLOR
689 Specifies whether to use color in diagnostic messages. Possible
690 values are always, auto and never.
691 This utility, like most other PostgreSQL utilities, also uses the
693 environment variables supported by libpq (see Section 33.14).
694
696 pg_dump internally executes SELECT statements. If you have problems
697 running pg_dump, make sure you are able to select information from the
698 database using, for example, psql(1). Also, any default connection
699 settings and environment variables used by the libpq front-end library
700 will apply.
701 The database activity of pg_dump is normally collected by the
703 statistics collector. If this is undesirable, you can set parameter
704 track_counts to false via PGOPTIONS or the ALTER USER command.
705
707 If your database cluster has any local additions to the template1
708 database, be careful to restore the output of pg_dump into a truly
709 empty database; otherwise you are likely to get errors due to duplicate
710 definitions of the added objects. To make an empty database without any
711 local additions, copy from template0 not template1, for example:
712 CREATE DATABASE foo WITH TEMPLATE template0;
714 When a data-only dump is chosen and the option --disable-triggers is
716 used, pg_dump emits commands to disable triggers on user tables before
717 inserting the data, and then commands to re-enable them after the data
718 has been inserted. If the restore is stopped in the middle, the system
719 catalogs might be left in the wrong state.
720 The dump file produced by pg_dump does not contain the statistics used
722 by the optimizer to make query planning decisions. Therefore, it is
723 wise to run ANALYZE after restoring from a dump file to ensure optimal
724 performance; see Section 24.1.3 and Section 24.1.6 for more
725 information.
726 Because pg_dump is used to transfer data to newer versions of
728 PostgreSQL, the output of pg_dump can be expected to load into
729 PostgreSQL server versions newer than pg_dump's version. pg_dump can
730 also dump from PostgreSQL servers older than its own version.
731 (Currently, servers back to version 8.0 are supported.) However,
732 pg_dump cannot dump from PostgreSQL servers newer than its own major
733 version; it will refuse to even try, rather than risk making an invalid
734 dump. Also, it is not guaranteed that pg_dump's output can be loaded
735 into a server of an older major version — not even if the dump was
736 taken from a server of that version. Loading a dump file into an older
737 server may require manual editing of the dump file to remove syntax not
738 understood by the older server. Use of the --quote-all-identifiers
739 option is recommended in cross-version cases, as it can prevent
740 problems arising from varying reserved-word lists in different
741 PostgreSQL versions.
742 When dumping logical replication subscriptions, pg_dump will generate
744 CREATE SUBSCRIPTION commands that use the connect = false option, so
745 that restoring the subscription does not make remote connections for
746 creating a replication slot or for initial table copy. That way, the
747 dump can be restored without requiring network access to the remote
748 servers. It is then up to the user to reactivate the subscriptions in a
749 suitable way. If the involved hosts have changed, the connection
750 information might have to be changed. It might also be appropriate to
751 truncate the target tables before initiating a new full table copy.
752
754 To dump a database called mydb into a SQL-script file:
755 $ pg_dump mydb > db.sql
757 To reload such a script into a (freshly created) database named newdb:
759 $ psql -d newdb -f db.sql
761 To dump a database into a custom-format archive file:
763 $ pg_dump -Fc mydb > db.dump
765 To dump a database into a directory-format archive:
767 $ pg_dump -Fd mydb -f dumpdir
769 To dump a database into a directory-format archive in parallel with 5
771 worker jobs:
772 $ pg_dump -Fd mydb -j 5 -f dumpdir
774 To reload an archive file into a (freshly created) database named
776 newdb:
777 $ pg_restore -d newdb db.dump
779 To reload an archive file into the same database it was dumped from,
781 discarding the current contents of that database:
782 $ pg_restore -d postgres --clean --create db.dump
784 To dump a single table named mytab:
786 $ pg_dump -t mytab mydb > db.sql
788 To dump all tables whose names start with emp in the detroit schema,
790 except for the table named employee_log:
791 $ pg_dump -t 'detroit.emp*' -T detroit.employee_log mydb > db.sql
793 To dump all schemas whose names start with east or west and end in gsm,
795 excluding any schemas whose names contain the word test:
796 $ pg_dump -n 'east*gsm' -n 'west*gsm' -N '*test*' mydb > db.sql
798 The same, using regular expression notation to consolidate the
800 switches:
801 $ pg_dump -n '(east|west)*gsm' -N '*test*' mydb > db.sql
803 To dump all database objects except for tables whose names begin with
805 ts_:
806 $ pg_dump -T 'ts_*' mydb > db.sql
808 To specify an upper-case or mixed-case name in -t and related switches,
810 you need to double-quote the name; else it will be folded to lower case
811 (see Patterns below). But double quotes are special to the shell, so in
812 turn they must be quoted. Thus, to dump a single table with a
813 mixed-case name, you need something like
814 $ pg_dump -t "\"MixedCaseName\"" mydb > mytab.sql
816
818 pg_dumpall(1), pg_restore(1), psql(1)
819PostgreSQL 13.4 2021 PG_DUMP(1)