Baoding-FenShuaJiang
/
Green_GoodERP18_FSJ_Standard
Watch 1
Star 0
Wiki 捐赠
gooderp18绿色标准版
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Tree: 0f75e31513
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from '0f75e31513'
${ noResults }
Green_GoodERP18_FSJ_Standard/PostgreSQL/doc/postgresql/html/app-pg-dumpall.html

350 lines
30KB
Raw Blame History 

  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>pg_dumpall</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 V1.79.1" /><link rel="prev" href="app-pgdump.html" title="pg_dump" /><link rel="next" href="app-pg-isready.html" title="pg_isready" /></head><body><div xmlns="http://www.w3.org/TR/xhtml1/transitional" class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center"><span xmlns="http://www.w3.org/1999/xhtml" class="application">pg_dumpall</span></th></tr><tr><td width="10%" align="left"><a accesskey="p" href="app-pgdump.html" title="pg_dump">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 12.4 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="app-pg-isready.html" title="pg_isready">Next</a></td></tr></table><hr></hr></div><div class="refentry" id="APP-PG-DUMPALL"><div class="titlepage"></div><a id="id-1.9.4.13.1" class="indexterm"></a><div class="refnamediv"><h2><span class="refentrytitle"><span class="application">pg_dumpall</span></span></h2><p>pg_dumpall — extract a <span class="productname">PostgreSQL</span> database cluster into a script file</p></div><div class="refsynopsisdiv"><h2>Synopsis</h2><div class="cmdsynopsis"><p id="id-1.9.4.13.4.1"><code class="command">pg_dumpall</code> [<em class="replaceable"><code>connection-option</code></em>...] [<em class="replaceable"><code>option</code></em>...]</p></div></div><div class="refsect1" id="APP-PG-DUMPALL-DESCRIPTION"><h2>Description</h2><p>

  3.    <span class="application">pg_dumpall</span> is a utility for writing out

  4.    (<span class="quote">“<span class="quote">dumping</span>”</span>) all <span class="productname">PostgreSQL</span> databases

  5.    of a cluster into one script file.  The script file contains

  6.    <acronym class="acronym">SQL</acronym> commands that can be used as input to <a class="xref" href="app-psql.html" title="psql"><span class="refentrytitle"><span class="application">psql</span></span></a> to restore the databases.  It does this by

  7.    calling <a class="xref" href="app-pgdump.html" title="pg_dump"><span class="refentrytitle">pg_dump</span></a> for each database in the cluster.

  8.    <span class="application">pg_dumpall</span> also dumps global objects

  9.    that are common to all databases, that is, database roles and tablespaces.

  10.    (<span class="application">pg_dump</span> does not save these objects.)

  11.   </p><p>

  12.    Since <span class="application">pg_dumpall</span> reads tables from all

  13.    databases you will most likely have to connect as a database

  14.    superuser in order to produce a complete dump.  Also you will need

  15.    superuser privileges to execute the saved script in order to be

  16.    allowed to add roles and create databases.

  17.   </p><p>

  18.    The SQL script will be written to the standard output.  Use the

  19.    <code class="option">-f</code>/<code class="option">--file</code> option or shell operators to

  20.    redirect it into a file.

  21.   </p><p>

  22.   <span class="application">pg_dumpall</span> needs to connect several

  23.   times to the <span class="productname">PostgreSQL</span> server (once per

  24.   database).  If you use password authentication it will ask for

  25.   a password each time. It is convenient to have a

  26.   <code class="filename">~/.pgpass</code> file in such cases. See <a class="xref" href="libpq-pgpass.html" title="33.15. The Password File">Section 33.15</a> for more information.

  27.   </p></div><div class="refsect1" id="id-1.9.4.13.6"><h2>Options</h2><p>

  28.     The following command-line options control the content and

  29.     format of the output.

  30. 

  31.     </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="option">-a</code><br /></span><span class="term"><code class="option">--data-only</code></span></dt><dd><p>

  32.         Dump only the data, not the schema (data definitions).

  33.        </p></dd><dt><span class="term"><code class="option">-c</code><br /></span><span class="term"><code class="option">--clean</code></span></dt><dd><p>

  34.         Include SQL commands to clean (drop) databases before

  35.         recreating them.  <code class="command">DROP</code> commands for roles and

  36.         tablespaces are added as well.

  37.        </p></dd><dt><span class="term"><code class="option">-E <em class="replaceable"><code>encoding</code></em></code><br /></span><span class="term"><code class="option">--encoding=<em class="replaceable"><code>encoding</code></em></code></span></dt><dd><p>

  38.         Create the dump in the specified character set encoding. By default,

  39.         the dump is created in the database encoding.  (Another way to get the

  40.         same result is to set the <code class="envar">PGCLIENTENCODING</code> environment

  41.         variable to the desired dump encoding.)

  42.        </p></dd><dt><span class="term"><code class="option">-f <em class="replaceable"><code>filename</code></em></code><br /></span><span class="term"><code class="option">--file=<em class="replaceable"><code>filename</code></em></code></span></dt><dd><p>

  43.         Send output to the specified file.  If this is omitted, the

  44.         standard output is used.

  45.        </p></dd><dt><span class="term"><code class="option">-g</code><br /></span><span class="term"><code class="option">--globals-only</code></span></dt><dd><p>

  46.         Dump only global objects (roles and tablespaces), no databases.

  47.        </p></dd><dt><span class="term"><code class="option">-O</code><br /></span><span class="term"><code class="option">--no-owner</code></span></dt><dd><p>

  48.         Do not output commands to set

  49.         ownership of objects to match the original database.

  50.         By default, <span class="application">pg_dumpall</span> issues

  51.         <code class="command">ALTER OWNER</code> or

  52.         <code class="command">SET SESSION AUTHORIZATION</code>

  53.         statements to set ownership of created schema elements.

  54.         These statements

  55.         will fail when the script is run unless it is started by a superuser

  56.         (or the same user that owns all of the objects in the script).

  57.         To make a script that can be restored by any user, but will give

  58.         that user ownership of all the objects, specify <code class="option">-O</code>.

  59.        </p></dd><dt><span class="term"><code class="option">-r</code><br /></span><span class="term"><code class="option">--roles-only</code></span></dt><dd><p>

  60.         Dump only roles, no databases or tablespaces.

  61.        </p></dd><dt><span class="term"><code class="option">-s</code><br /></span><span class="term"><code class="option">--schema-only</code></span></dt><dd><p>

  62.         Dump only the object definitions (schema), not data.

  63.        </p></dd><dt><span class="term"><code class="option">-S <em class="replaceable"><code>username</code></em></code><br /></span><span class="term"><code class="option">--superuser=<em class="replaceable"><code>username</code></em></code></span></dt><dd><p>

  64.         Specify the superuser user name to use when disabling triggers.

  65.         This is relevant only if <code class="option">--disable-triggers</code> is used.

  66.         (Usually, it's better to leave this out, and instead start the

  67.         resulting script as superuser.)

  68.        </p></dd><dt><span class="term"><code class="option">-t</code><br /></span><span class="term"><code class="option">--tablespaces-only</code></span></dt><dd><p>

  69.         Dump only tablespaces, no databases or roles.

  70.        </p></dd><dt><span class="term"><code class="option">-v</code><br /></span><span class="term"><code class="option">--verbose</code></span></dt><dd><p>

  71.         Specifies verbose mode.  This will cause

  72.         <span class="application">pg_dumpall</span> to output start/stop

  73.         times to the dump file, and progress messages to standard error.

  74.         It will also enable verbose output in <span class="application">pg_dump</span>.

  75.        </p></dd><dt><span class="term"><code class="option">-V</code><br /></span><span class="term"><code class="option">--version</code></span></dt><dd><p>

  76.        Print the <span class="application">pg_dumpall</span> version and exit.

  77.        </p></dd><dt><span class="term"><code class="option">-x</code><br /></span><span class="term"><code class="option">--no-privileges</code><br /></span><span class="term"><code class="option">--no-acl</code></span></dt><dd><p>

  78.         Prevent dumping of access privileges (grant/revoke commands).

  79.        </p></dd><dt><span class="term"><code class="option">--binary-upgrade</code></span></dt><dd><p>

  80.         This option is for use by in-place upgrade utilities.  Its use

  81.         for other purposes is not recommended or supported.  The

  82.         behavior of the option may change in future releases without

  83.         notice.

  84.        </p></dd><dt><span class="term"><code class="option">--column-inserts</code><br /></span><span class="term"><code class="option">--attribute-inserts</code></span></dt><dd><p>

  85.         Dump data as <code class="command">INSERT</code> commands with explicit

  86.         column names (<code class="literal">INSERT INTO

  87.         <em class="replaceable"><code>table</code></em>

  88.         (<em class="replaceable"><code>column</code></em>, ...) VALUES

  89.         ...</code>).  This will make restoration very slow; it is mainly

  90.         useful for making dumps that can be loaded into

  91.         non-<span class="productname">PostgreSQL</span> databases.

  92.        </p></dd><dt><span class="term"><code class="option">--disable-dollar-quoting</code></span></dt><dd><p>

  93.         This option disables the use of dollar quoting for function bodies,

  94.         and forces them to be quoted using SQL standard string syntax.

  95.        </p></dd><dt><span class="term"><code class="option">--disable-triggers</code></span></dt><dd><p>

  96.         This option is relevant only when creating a data-only dump.

  97.         It instructs <span class="application">pg_dumpall</span> to include commands

  98.         to temporarily disable triggers on the target tables while

  99.         the data is reloaded.  Use this if you have referential

  100.         integrity checks or other triggers on the tables that you

  101.         do not want to invoke during data reload.

  102.        </p><p>

  103.         Presently, the commands emitted for <code class="option">--disable-triggers</code>

  104.         must be done as superuser.  So, you should also specify

  105.         a superuser name with <code class="option">-S</code>, or preferably be careful to

  106.         start the resulting script as a superuser.

  107.        </p></dd><dt><span class="term"><code class="option">--extra-float-digits=<em class="replaceable"><code>ndigits</code></em></code></span></dt><dd><p>

  108.         Use the specified value of extra_float_digits when dumping

  109.         floating-point data, instead of the maximum available precision.

  110.         Routine dumps made for backup purposes should not use this option.

  111.        </p></dd><dt><span class="term"><code class="option">--exclude-database=<em class="replaceable"><code>pattern</code></em></code></span></dt><dd><p>

  112.         Do not dump databases whose name matches

  113.         <em class="replaceable"><code>pattern</code></em>.

  114.         Multiple patterns can be excluded by writing multiple

  115.         <code class="option">--exclude-database</code> switches.  The

  116.         <em class="replaceable"><code>pattern</code></em> parameter is

  117.         interpreted as a pattern according to the same rules used by

  118.         <span class="application">psql</span>'s <code class="literal">\d</code>

  119.         commands (see <a class="xref" href="app-psql.html#APP-PSQL-PATTERNS" title="Patterns">Patterns</a>),

  120.         so multiple databases can also be excluded by writing wildcard

  121.         characters in the pattern.  When using wildcards, be careful to

  122.         quote the pattern if needed to prevent shell wildcard expansion.

  123.        </p></dd><dt><span class="term"><code class="option">--if-exists</code></span></dt><dd><p>

  124.         Use conditional commands (i.e. add an <code class="literal">IF EXISTS</code>

  125.         clause) to drop databases and other objects.  This option is not valid

  126.         unless <code class="option">--clean</code> is also specified.

  127.        </p></dd><dt><span class="term"><code class="option">--inserts</code></span></dt><dd><p>

  128.         Dump data as <code class="command">INSERT</code> commands (rather

  129.         than <code class="command">COPY</code>).  This will make restoration very slow;

  130.         it is mainly useful for making dumps that can be loaded into

  131.         non-<span class="productname">PostgreSQL</span> databases.  Note that

  132.         the restore might fail altogether if you have rearranged column order.

  133.         The <code class="option">--column-inserts</code> option is safer, though even

  134.         slower.

  135.        </p></dd><dt><span class="term"><code class="option">--load-via-partition-root</code></span></dt><dd><p>

  136.         When dumping data for a table partition, make

  137.         the <code class="command">COPY</code> or <code class="command">INSERT</code> statements

  138.         target the root of the partitioning hierarchy that contains it, rather

  139.         than the partition itself.  This causes the appropriate partition to

  140.         be re-determined for each row when the data is loaded.  This may be

  141.         useful when reloading data on a server where rows do not always fall

  142.         into the same partitions as they did on the original server.  That

  143.         could happen, for example, if the partitioning column is of type text

  144.         and the two systems have different definitions of the collation used

  145.         to sort the partitioning column.

  146.        </p></dd><dt><span class="term"><code class="option">--lock-wait-timeout=<em class="replaceable"><code>timeout</code></em></code></span></dt><dd><p>

  147.         Do not wait forever to acquire shared table locks at the beginning of

  148.         the dump. Instead, fail if unable to lock a table within the specified

  149.         <em class="replaceable"><code>timeout</code></em>. The timeout may be

  150.         specified in any of the formats accepted by <code class="command">SET

  151.         statement_timeout</code>.  Allowed values vary depending on the server

  152.         version you are dumping from, but an integer number of milliseconds

  153.         is accepted by all versions since 7.3.  This option is ignored when

  154.         dumping from a pre-7.3 server.

  155.        </p></dd><dt><span class="term"><code class="option">--no-comments</code></span></dt><dd><p>

  156.         Do not dump comments.

  157.        </p></dd><dt><span class="term"><code class="option">--no-publications</code></span></dt><dd><p>

  158.         Do not dump publications.

  159.        </p></dd><dt><span class="term"><code class="option">--no-role-passwords</code></span></dt><dd><p>

  160.         Do not dump passwords for roles.  When restored, roles will have a

  161.         null password, and password authentication will always fail until the

  162.         password is set.  Since password values aren't needed when this option

  163.         is specified, the role information is read from the catalog

  164.         view <code class="structname">pg_roles</code> instead

  165.         of <code class="structname">pg_authid</code>.  Therefore, this option also

  166.         helps if access to <code class="structname">pg_authid</code> is restricted by

  167.         some security policy.

  168.        </p></dd><dt><span class="term"><code class="option">--no-security-labels</code></span></dt><dd><p>

  169.         Do not dump security labels.

  170.        </p></dd><dt><span class="term"><code class="option">--no-subscriptions</code></span></dt><dd><p>

  171.         Do not dump subscriptions.

  172.        </p></dd><dt><span class="term"><code class="option">--no-sync</code></span></dt><dd><p>

  173.         By default, <code class="command">pg_dumpall</code> will wait for all files

  174.         to be written safely to disk.  This option causes

  175.         <code class="command">pg_dumpall</code> to return without waiting, which is

  176.         faster, but means that a subsequent operating system crash can leave

  177.         the dump corrupt.  Generally, this option is useful for testing

  178.         but should not be used when dumping data from production installation.

  179.        </p></dd><dt><span class="term"><code class="option">--no-tablespaces</code></span></dt><dd><p>

  180.         Do not output commands to create tablespaces nor select tablespaces

  181.         for objects.

  182.         With this option, all objects will be created in whichever

  183.         tablespace is the default during restore.

  184.        </p></dd><dt><span class="term"><code class="option">--no-unlogged-table-data</code></span></dt><dd><p>

  185.         Do not dump the contents of unlogged tables.  This option has no

  186.         effect on whether or not the table definitions (schema) are dumped;

  187.         it only suppresses dumping the table data.

  188.        </p></dd><dt><span class="term"><code class="option">--on-conflict-do-nothing</code></span></dt><dd><p>

  189.         Add <code class="literal">ON CONFLICT DO NOTHING</code> to

  190.         <code class="command">INSERT</code> commands.

  191.         This option is not valid unless <code class="option">--inserts</code> or

  192.         <code class="option">--column-inserts</code> is also specified.

  193.        </p></dd><dt><span class="term"><code class="option">--quote-all-identifiers</code></span></dt><dd><p>

  194.         Force quoting of all identifiers.  This option is recommended when

  195.         dumping a database from a server whose <span class="productname">PostgreSQL</span>

  196.         major version is different from <span class="application">pg_dumpall</span>'s, or when

  197.         the output is intended to be loaded into a server of a different

  198.         major version.  By default, <span class="application">pg_dumpall</span> quotes only

  199.         identifiers that are reserved words in its own major version.

  200.         This sometimes results in compatibility issues when dealing with

  201.         servers of other versions that may have slightly different sets

  202.         of reserved words.  Using <code class="option">--quote-all-identifiers</code> prevents

  203.         such issues, at the price of a harder-to-read dump script.

  204.        </p></dd><dt><span class="term"><code class="option">--rows-per-insert=<em class="replaceable"><code>nrows</code></em></code></span></dt><dd><p>

  205.         Dump data as <code class="command">INSERT</code> commands (rather than

  206.         <code class="command">COPY</code>).  Controls the maximum number of rows per

  207.         <code class="command">INSERT</code> command. The value specified must be a

  208.         number greater than zero.  Any error during reloading will cause only

  209.         rows that are part of the problematic <code class="command">INSERT</code> to be

  210.         lost, rather than the entire table contents.

  211.        </p></dd><dt><span class="term"><code class="option">--use-set-session-authorization</code></span></dt><dd><p>

  212.         Output SQL-standard <code class="command">SET SESSION AUTHORIZATION</code> commands

  213.         instead of <code class="command">ALTER OWNER</code> commands to determine object

  214.         ownership.  This makes the dump more standards compatible, but

  215.         depending on the history of the objects in the dump, might not restore

  216.         properly.

  217.        </p></dd><dt><span class="term"><code class="option">-?</code><br /></span><span class="term"><code class="option">--help</code></span></dt><dd><p>

  218.        Show help about <span class="application">pg_dumpall</span> command line

  219.        arguments, and exit.

  220.        </p></dd></dl></div><p>

  221.    </p><p>

  222.    The following command-line options control the database connection parameters.

  223. 

  224.    </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="option">-d <em class="replaceable"><code>connstr</code></em></code><br /></span><span class="term"><code class="option">--dbname=<em class="replaceable"><code>connstr</code></em></code></span></dt><dd><p>

  225.         Specifies parameters used to connect to the server, as a connection

  226.         string. See <a class="xref" href="libpq-connect.html#LIBPQ-CONNSTRING" title="33.1.1. Connection Strings">Section 33.1.1</a> for more information.

  227.        </p><p>

  228.         The option is called <code class="literal">--dbname</code> for consistency with other

  229.         client applications, but because <span class="application">pg_dumpall</span>

  230.         needs to connect to many databases, the database name in the

  231.         connection string will be ignored.  Use the <code class="literal">-l</code>

  232.         option to specify the name of the database used for the initial

  233.         connection, which will dump global objects and discover what other

  234.         databases should be dumped.

  235.        </p></dd><dt><span class="term"><code class="option">-h <em class="replaceable"><code>host</code></em></code><br /></span><span class="term"><code class="option">--host=<em class="replaceable"><code>host</code></em></code></span></dt><dd><p>

  236.         Specifies the host name of the machine on which the database

  237.         server is running.  If the value begins with a slash, it is

  238.         used as the directory for the Unix domain socket.  The default

  239.         is taken from the <code class="envar">PGHOST</code> environment variable,

  240.         if set, else a Unix domain socket connection is attempted.

  241.        </p></dd><dt><span class="term"><code class="option">-l <em class="replaceable"><code>dbname</code></em></code><br /></span><span class="term"><code class="option">--database=<em class="replaceable"><code>dbname</code></em></code></span></dt><dd><p>

  242.          Specifies the name of the database to connect to for dumping global

  243.          objects and discovering what other databases should be dumped. If

  244.          not specified, the <code class="literal">postgres</code> database will be used,

  245.          and if that does not exist, <code class="literal">template1</code> will be used.

  246.        </p></dd><dt><span class="term"><code class="option">-p <em class="replaceable"><code>port</code></em></code><br /></span><span class="term"><code class="option">--port=<em class="replaceable"><code>port</code></em></code></span></dt><dd><p>

  247.         Specifies the TCP port or local Unix domain socket file

  248.         extension on which the server is listening for connections.

  249.         Defaults to the <code class="envar">PGPORT</code> environment variable, if

  250.         set, or a compiled-in default.

  251.        </p></dd><dt><span class="term"><code class="option">-U <em class="replaceable"><code>username</code></em></code><br /></span><span class="term"><code class="option">--username=<em class="replaceable"><code>username</code></em></code></span></dt><dd><p>

  252.         User name to connect as.

  253.        </p></dd><dt><span class="term"><code class="option">-w</code><br /></span><span class="term"><code class="option">--no-password</code></span></dt><dd><p>

  254.         Never issue a password prompt.  If the server requires

  255.         password authentication and a password is not available by

  256.         other means such as a <code class="filename">.pgpass</code> file, the

  257.         connection attempt will fail.  This option can be useful in

  258.         batch jobs and scripts where no user is present to enter a

  259.         password.

  260.        </p></dd><dt><span class="term"><code class="option">-W</code><br /></span><span class="term"><code class="option">--password</code></span></dt><dd><p>

  261.         Force <span class="application">pg_dumpall</span> to prompt for a

  262.         password before connecting to a database.

  263.        </p><p>

  264.         This option is never essential, since

  265.         <span class="application">pg_dumpall</span> will automatically prompt

  266.         for a password if the server demands password authentication.

  267.         However, <span class="application">pg_dumpall</span> will waste a

  268.         connection attempt finding out that the server wants a password.

  269.         In some cases it is worth typing <code class="option">-W</code> to avoid the extra

  270.         connection attempt.

  271.        </p><p>

  272.         Note that the password prompt will occur again for each database

  273.         to be dumped.  Usually, it's better to set up a

  274.         <code class="filename">~/.pgpass</code> file than to rely on manual password entry.

  275.        </p></dd><dt><span class="term"><code class="option">--role=<em class="replaceable"><code>rolename</code></em></code></span></dt><dd><p>

  276.         Specifies a role name to be used to create the dump.

  277.         This option causes <span class="application">pg_dumpall</span> to issue a

  278.         <code class="command">SET ROLE</code> <em class="replaceable"><code>rolename</code></em>

  279.         command after connecting to the database. It is useful when the

  280.         authenticated user (specified by <code class="option">-U</code>) lacks privileges

  281.         needed by <span class="application">pg_dumpall</span>, but can switch to a role with

  282.         the required rights.  Some installations have a policy against

  283.         logging in directly as a superuser, and use of this option allows

  284.         dumps to be made without violating the policy.

  285.        </p></dd></dl></div><p>

  286.   </p></div><div class="refsect1" id="id-1.9.4.13.7"><h2>Environment</h2><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="envar">PGHOST</code><br /></span><span class="term"><code class="envar">PGOPTIONS</code><br /></span><span class="term"><code class="envar">PGPORT</code><br /></span><span class="term"><code class="envar">PGUSER</code></span></dt><dd><p>

  287.       Default connection parameters

  288.      </p></dd><dt><span class="term"><code class="envar">PG_COLOR</code></span></dt><dd><p>

  289.       Specifies whether to use color in diagnostic messages. Possible values

  290.       are <code class="literal">always</code>, <code class="literal">auto</code> and

  291.       <code class="literal">never</code>.

  292.      </p></dd></dl></div><p>

  293.    This utility, like most other <span class="productname">PostgreSQL</span> utilities,

  294.    also uses the environment variables supported by <span class="application">libpq</span>

  295.    (see <a class="xref" href="libpq-envars.html" title="33.14. Environment Variables">Section 33.14</a>).

  296.   </p></div><div class="refsect1" id="id-1.9.4.13.8"><h2>Notes</h2><p>

  297.    Since <span class="application">pg_dumpall</span> calls

  298.    <span class="application">pg_dump</span> internally, some diagnostic

  299.    messages will refer to <span class="application">pg_dump</span>.

  300.   </p><p>

  301.    The <code class="option">--clean</code> option can be useful even when your

  302.    intention is to restore the dump script into a fresh cluster.  Use of

  303.    <code class="option">--clean</code> authorizes the script to drop and re-create the

  304.    built-in <code class="literal">postgres</code> and <code class="literal">template1</code>

  305.    databases, ensuring that those databases will retain the same properties

  306.    (for instance, locale and encoding) that they had in the source cluster.

  307.    Without the option, those databases will retain their existing

  308.    database-level properties, as well as any pre-existing contents.

  309.   </p><p>

  310.    Once restored, it is wise to run <code class="command">ANALYZE</code> on each

  311.    database so the optimizer has useful statistics. You

  312.    can also run <code class="command">vacuumdb -a -z</code> to analyze all

  313.    databases.

  314.   </p><p>

  315.    The dump script should not be expected to run completely without errors.

  316.    In particular, because the script will issue <code class="command">CREATE ROLE</code>

  317.    for every role existing in the source cluster, it is certain to get a

  318.    <span class="quote">“<span class="quote">role already exists</span>”</span> error for the bootstrap superuser,

  319.    unless the destination cluster was initialized with a different bootstrap

  320.    superuser name.  This error is harmless and should be ignored.  Use of

  321.    the <code class="option">--clean</code> option is likely to produce additional

  322.    harmless error messages about non-existent objects, although you can

  323.    minimize those by adding <code class="option">--if-exists</code>.

  324.   </p><p>

  325.    <span class="application">pg_dumpall</span> requires all needed

  326.    tablespace directories to exist before the restore;  otherwise,

  327.    database creation will fail for databases in non-default

  328.    locations.

  329.   </p></div><div class="refsect1" id="APP-PG-DUMPALL-EX"><h2>Examples</h2><p>

  330.    To dump all databases:

  331. 

  332. </p><pre class="screen">

  333. <code class="prompt">$</code> <strong class="userinput"><code>pg_dumpall &gt; db.out</code></strong>

  334. </pre><p>

  335.   </p><p>

  336.    To reload database(s) from this file, you can use:

  337. </p><pre class="screen">

  338. <code class="prompt">$</code> <strong class="userinput"><code>psql -f db.out postgres</code></strong>

  339. </pre><p>

  340.    It is not important to which database you connect here since the

  341.    script file created by <span class="application">pg_dumpall</span> will

  342.    contain the appropriate commands to create and connect to the saved

  343.    databases.  An exception is that if you specified <code class="option">--clean</code>,

  344.    you must connect to the <code class="literal">postgres</code> database initially;

  345.    the script will attempt to drop other databases immediately, and that

  346.    will fail for the database you are connected to.

  347.   </p></div><div class="refsect1" id="id-1.9.4.13.10"><h2>See Also</h2><p>

  348.     Check <a class="xref" href="app-pgdump.html" title="pg_dump"><span class="refentrytitle">pg_dump</span></a> for details on possible

  349.     error conditions.

  350.   </p></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="app-pgdump.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="reference-client.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="app-pg-isready.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">pg_dump </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> <span class="application">pg_isready</span></td></tr></table></div></body></html>
上海开阖软件有限公司 沪ICP备12045867号-1