Baoding-FenShuaJiang
/
Green_GoodERP18_FSJ_Standard
關註 1
收藏 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.
分支: master
分支列表 標籤列表
${ item.name }
Create branch ${ searchTerm }
from 'master'
${ noResults }
Green_GoodERP18_FSJ_Standard/PostgreSQL/doc/postgresql/html/app-pgbasebackup.html

406 line
33KB
原始文件 永久連結 Blame 文件歷史 

  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_basebackup</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-ecpg.html" title="ecpg" /><link rel="next" href="pgbench.html" title="pgbench" /></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">pg_basebackup</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="app-ecpg.html" title="ecpg">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="pgbench.html" title="pgbench">Next</a></td></tr></table><hr></hr></div><div class="refentry" id="APP-PGBASEBACKUP"><div class="titlepage"></div><a id="id-1.9.4.9.1" class="indexterm"></a><div class="refnamediv"><h2><span class="refentrytitle">pg_basebackup</span></h2><p>pg_basebackup — take a base backup of a <span class="productname">PostgreSQL</span> cluster</p></div><div class="refsynopsisdiv"><h2>Synopsis</h2><div class="cmdsynopsis"><p id="id-1.9.4.9.4.1"><code class="command">pg_basebackup</code> [<em class="replaceable"><code>option</code></em>...]</p></div></div><div class="refsect1" id="id-1.9.4.9.5"><h2>

  3.    Description

  4.   </h2><p>

  5.    <span class="application">pg_basebackup</span> is used to take base backups of

  6.    a running <span class="productname">PostgreSQL</span> database cluster. These

  7.    are taken without affecting other clients to the database, and can be used

  8.    both for point-in-time recovery (see <a class="xref" href="continuous-archiving.html" title="25.3. Continuous Archiving and Point-in-Time Recovery (PITR)">Section 25.3</a>)

  9.    and as the starting point for a log shipping or streaming replication standby

  10.    servers (see <a class="xref" href="warm-standby.html" title="26.2. Log-Shipping Standby Servers">Section 26.2</a>).

  11.   </p><p>

  12.    <span class="application">pg_basebackup</span> makes a binary copy of the database

  13.    cluster files, while making sure the system is put in and

  14.    out of backup mode automatically. Backups are always taken of the entire

  15.    database cluster; it is not possible to back up individual databases or

  16.    database objects. For individual database backups, a tool such as

  17.    <a class="xref" href="app-pgdump.html" title="pg_dump"><span class="refentrytitle">pg_dump</span></a> must be used.

  18.   </p><p>

  19.    The backup is made over a regular <span class="productname">PostgreSQL</span>

  20.    connection, and uses the replication protocol. The connection must be made

  21.    with a superuser or a user having <code class="literal">REPLICATION</code>

  22.    permissions (see <a class="xref" href="role-attributes.html" title="21.2. Role Attributes">Section 21.2</a>),

  23.    and <code class="filename">pg_hba.conf</code> must explicitly permit the replication

  24.    connection. The server must also be configured

  25.    with <a class="xref" href="runtime-config-replication.html#GUC-MAX-WAL-SENDERS">max_wal_senders</a> set high enough to leave at least

  26.    one session available for the backup and one for WAL streaming (if used).

  27.   </p><p>

  28.    There can be multiple <code class="command">pg_basebackup</code>s running at the same time, but it is

  29.    better from a performance point of view to take only one backup, and copy

  30.    the result.

  31.   </p><p>

  32.    <span class="application">pg_basebackup</span> can make a base backup from

  33.    not only the master but also the standby. To take a backup from the standby,

  34.    set up the standby so that it can accept replication connections (that is, set

  35.    <code class="varname">max_wal_senders</code> and <a class="xref" href="runtime-config-replication.html#GUC-HOT-STANDBY">hot_standby</a>,

  36.    and configure <a class="link" href="auth-pg-hba-conf.html" title="20.1. The pg_hba.conf File">host-based authentication</a>).

  37.    You will also need to enable <a class="xref" href="runtime-config-wal.html#GUC-FULL-PAGE-WRITES">full_page_writes</a> on the master.

  38.   </p><p>

  39.    Note that there are some limitations in an online backup from the standby:

  40. 

  41.    </p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>

  42.       The backup history file is not created in the database cluster backed up.

  43.      </p></li><li class="listitem"><p>

  44.       If you are using <code class="literal">-X none</code>, there is no guarantee that all

  45.       WAL files required for the backup are archived at the end of backup.

  46.      </p></li><li class="listitem"><p>

  47.       If the standby is promoted to the master during online backup, the backup fails.

  48.      </p></li><li class="listitem"><p>

  49.       All WAL records required for the backup must contain sufficient full-page writes,

  50.       which requires you to enable <code class="varname">full_page_writes</code> on the master and

  51.       not to use a tool like <span class="application">pg_compresslog</span> as

  52.       <code class="varname">archive_command</code> to remove full-page writes from WAL files.

  53.      </p></li></ul></div><p>

  54.   </p></div><div class="refsect1" id="id-1.9.4.9.6"><h2>Options</h2><p>

  55.     The following command-line options control the location and format of the

  56.     output.

  57. 

  58.     </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="option">-D <em class="replaceable"><code>directory</code></em></code><br /></span><span class="term"><code class="option">--pgdata=<em class="replaceable"><code>directory</code></em></code></span></dt><dd><p>

  59.         Directory to write the output to.

  60.         <span class="application">pg_basebackup</span> will create the directory and

  61.         any parent directories if necessary.  The directory may already exist,

  62.         but it is an error if the directory already exists and is not empty.

  63.        </p><p>

  64.         When the backup is in tar mode, and the directory is specified as

  65.         <code class="literal">-</code> (dash), the tar file will be written to

  66.         <code class="literal">stdout</code>.

  67.        </p><p>

  68.         This option is required.

  69.        </p></dd><dt><span class="term"><code class="option">-F <em class="replaceable"><code>format</code></em></code><br /></span><span class="term"><code class="option">--format=<em class="replaceable"><code>format</code></em></code></span></dt><dd><p>

  70.         Selects the format for the output. <em class="replaceable"><code>format</code></em>

  71.         can be one of the following:

  72. 

  73.         </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="literal">p</code><br /></span><span class="term"><code class="literal">plain</code></span></dt><dd><p>

  74.             Write the output as plain files, with the same layout as the

  75.             current data directory and tablespaces. When the cluster has

  76.             no additional tablespaces, the whole database will be placed in

  77.             the target directory. If the cluster contains additional

  78.             tablespaces, the main data directory will be placed in the

  79.             target directory, but all other tablespaces will be placed

  80.             in the same absolute path as they have on the server.

  81.            </p><p>

  82.             This is the default format.

  83.            </p></dd><dt><span class="term"><code class="literal">t</code><br /></span><span class="term"><code class="literal">tar</code></span></dt><dd><p>

  84.             Write the output as tar files in the target directory. The main

  85.             data directory will be written to a file named

  86.             <code class="filename">base.tar</code>, and all other tablespaces will

  87.             be named after the tablespace OID.

  88.             </p><p>

  89.             If the value <code class="literal">-</code> (dash) is specified as

  90.             target directory, the tar contents will be written to

  91.             standard output, suitable for piping to for example

  92.             <span class="productname">gzip</span>. This is only possible if

  93.             the cluster has no additional tablespaces and WAL

  94.             streaming is not used.

  95.            </p></dd></dl></div></dd><dt><span class="term"><code class="option">-r <em class="replaceable"><code>rate</code></em></code><br /></span><span class="term"><code class="option">--max-rate=<em class="replaceable"><code>rate</code></em></code></span></dt><dd><p>

  96.         The maximum transfer rate of data transferred from the server.  Values are

  97.         in kilobytes per second.  Use a suffix of <code class="literal">M</code> to indicate megabytes

  98.         per second.  A suffix of <code class="literal">k</code> is also accepted, and has no effect.

  99.         Valid values are between 32 kilobytes per second and 1024 megabytes per second.

  100.        </p><p>

  101.         The purpose is to limit the impact of <span class="application">pg_basebackup</span>

  102.         on the running server.

  103.        </p><p>

  104.         This option always affects transfer of the data directory. Transfer of

  105.         WAL files is only affected if the collection method is <code class="literal">fetch</code>.

  106.        </p></dd><dt><span class="term"><code class="option">-R</code><br /></span><span class="term"><code class="option">--write-recovery-conf</code></span></dt><dd><p>

  107.         Create <code class="filename">standby.signal</code> and append connection settings

  108.         to <code class="filename">postgresql.auto.conf</code> in the output

  109.         directory (or into the base archive file when using tar format) to

  110.         ease setting up a standby server.

  111.         The <code class="filename">postgresql.auto.conf</code> file will record the connection

  112.         settings and, if specified, the replication slot

  113.         that <span class="application">pg_basebackup</span> is using, so that the

  114.         streaming replication will use the same settings later on.

  115.        </p></dd><dt><span class="term"><code class="option">-T <em class="replaceable"><code>olddir</code></em>=<em class="replaceable"><code>newdir</code></em></code><br /></span><span class="term"><code class="option">--tablespace-mapping=<em class="replaceable"><code>olddir</code></em>=<em class="replaceable"><code>newdir</code></em></code></span></dt><dd><p>

  116.         Relocate the tablespace in directory <em class="replaceable"><code>olddir</code></em>

  117.         to <em class="replaceable"><code>newdir</code></em> during the backup.  To be

  118.         effective, <em class="replaceable"><code>olddir</code></em> must exactly match the

  119.         path specification of the tablespace as it is currently defined.  (But

  120.         it is not an error if there is no tablespace

  121.         in <em class="replaceable"><code>olddir</code></em> contained in the backup.)

  122.         Both <em class="replaceable"><code>olddir</code></em>

  123.         and <em class="replaceable"><code>newdir</code></em> must be absolute paths.  If a

  124.         path happens to contain a <code class="literal">=</code> sign, escape it with a

  125.         backslash.  This option can be specified multiple times for multiple

  126.         tablespaces.  See examples below.

  127.        </p><p>

  128.         If a tablespace is relocated in this way, the symbolic links inside

  129.         the main data directory are updated to point to the new location.  So

  130.         the new data directory is ready to be used for a new server instance

  131.         with all tablespaces in the updated locations.

  132.         </p></dd><dt><span class="term"><code class="option">--waldir=<em class="replaceable"><code>waldir</code></em></code></span></dt><dd><p>

  133.         Specifies the location for the write-ahead log directory.

  134.         <em class="replaceable"><code>waldir</code></em> must be an absolute path.

  135.         The write-ahead log directory can only be specified when

  136.         the backup is in plain mode.

  137.        </p></dd><dt><span class="term"><code class="option">-X <em class="replaceable"><code>method</code></em></code><br /></span><span class="term"><code class="option">--wal-method=<em class="replaceable"><code>method</code></em></code></span></dt><dd><p>

  138.         Includes the required write-ahead log files (WAL files) in the

  139.         backup. This will include all write-ahead logs generated during

  140.         the backup. Unless the method <code class="literal">none</code> is specified,

  141.         it is possible to start a postmaster directly in the extracted

  142.         directory without the need to consult the log archive, thus

  143.         making this a completely standalone backup.

  144.        </p><p>

  145.         The following methods for collecting the write-ahead logs are

  146.         supported:

  147. 

  148.         </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="literal">n</code><br /></span><span class="term"><code class="literal">none</code></span></dt><dd><p>

  149.             Don't include write-ahead log in the backup.

  150.            </p></dd><dt><span class="term"><code class="literal">f</code><br /></span><span class="term"><code class="literal">fetch</code></span></dt><dd><p>

  151.             The write-ahead log files are collected at the end of the backup.

  152.             Therefore, it is necessary for the

  153.             <a class="xref" href="runtime-config-replication.html#GUC-WAL-KEEP-SEGMENTS">wal_keep_segments</a> parameter to be set high

  154.              enough that the log is not removed before the end of the backup.

  155.              If the log has been rotated when it's time to transfer it, the

  156.              backup will fail and be unusable.

  157.            </p><p>

  158.             When tar format mode is used, the write-ahead log files will be

  159.             written to the <code class="filename">base.tar</code> file.

  160.            </p></dd><dt><span class="term"><code class="literal">s</code><br /></span><span class="term"><code class="literal">stream</code></span></dt><dd><p>

  161.             Stream the write-ahead log while the backup is created. This will

  162.             open a second connection to the server and start streaming the

  163.             write-ahead log in parallel while running the backup. Therefore,

  164.             it will use up two connections configured by the

  165.             <a class="xref" href="runtime-config-replication.html#GUC-MAX-WAL-SENDERS">max_wal_senders</a> parameter. As long as the

  166.              client can keep up with write-ahead log received, using this mode

  167.              requires no extra write-ahead logs to be saved on the master.

  168.            </p><p>

  169.             When tar format mode is used, the write-ahead log files will be

  170.             written to a separate file named <code class="filename">pg_wal.tar</code>

  171.             (if the server is a version earlier than 10, the file will be named

  172.             <code class="filename">pg_xlog.tar</code>).

  173.            </p><p>

  174.             This value is the default.

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

  176.        </p></dd><dt><span class="term"><code class="option">-z</code><br /></span><span class="term"><code class="option">--gzip</code></span></dt><dd><p>

  177.         Enables gzip compression of tar file output, with the default

  178.         compression level. Compression is only available when using

  179.         the tar format, and the suffix <code class="filename">.gz</code> will

  180.         automatically be added to all tar filenames.

  181.        </p></dd><dt><span class="term"><code class="option">-Z <em class="replaceable"><code>level</code></em></code><br /></span><span class="term"><code class="option">--compress=<em class="replaceable"><code>level</code></em></code></span></dt><dd><p>

  182.         Enables gzip compression of tar file output, and specifies the

  183.         compression level (0 through 9, 0 being no compression and 9 being best

  184.         compression). Compression is only available when using the tar

  185.         format, and the suffix <code class="filename">.gz</code> will

  186.         automatically be added to all tar filenames.

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

  188.    </p><p>

  189.     The following command-line options control the generation of the

  190.     backup and the running of the program.

  191. 

  192.     </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="option">-c <em class="replaceable"><code>fast|spread</code></em></code><br /></span><span class="term"><code class="option">--checkpoint=<em class="replaceable"><code>fast|spread</code></em></code></span></dt><dd><p>

  193.         Sets checkpoint mode to fast (immediate) or spread (default) (see <a class="xref" href="continuous-archiving.html#BACKUP-LOWLEVEL-BASE-BACKUP" title="25.3.3. Making a Base Backup Using the Low Level API">Section 25.3.3</a>).

  194.        </p></dd><dt><span class="term"><code class="option">-C</code><br /></span><span class="term"><code class="option">--create-slot</code></span></dt><dd><p>

  195.         This option causes creation of a replication slot named by the

  196.         <code class="literal">--slot</code> option before starting the backup.

  197.         An error is raised if the slot already exists.

  198.        </p></dd><dt><span class="term"><code class="option">-l <em class="replaceable"><code>label</code></em></code><br /></span><span class="term"><code class="option">--label=<em class="replaceable"><code>label</code></em></code></span></dt><dd><p>

  199.         Sets the label for the backup. If none is specified, a default value of

  200.         <span class="quote">“<span class="quote"><code class="literal">pg_basebackup base backup</code></span>”</span> will be used.

  201.        </p></dd><dt><span class="term"><code class="option">-n</code><br /></span><span class="term"><code class="option">--no-clean</code></span></dt><dd><p>

  202.         By default, when <code class="command">pg_basebackup</code> aborts with an

  203.         error, it removes any directories it might have created before

  204.         discovering that it cannot finish the job (for example, data directory

  205.         and write-ahead log directory). This option inhibits tidying-up and is

  206.         thus useful for debugging.

  207.        </p><p>

  208.         Note that tablespace directories are not cleaned up either way.

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

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

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

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

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

  214.         the base backup corrupt.  Generally, this option is useful for testing

  215.         but should not be used when creating a production installation.

  216.        </p></dd><dt><span class="term"><code class="option">-P</code><br /></span><span class="term"><code class="option">--progress</code></span></dt><dd><p>

  217.         Enables progress reporting. Turning this on will deliver an approximate

  218.         progress report during the backup. Since the database may change during

  219.         the backup, this is only an approximation and may not end at exactly

  220.         <code class="literal">100%</code>. In particular, when WAL log is included in the

  221.         backup, the total amount of data cannot be estimated in advance, and

  222.         in this case the estimated target size will increase once it passes the

  223.         total estimate without WAL.

  224.        </p><p>

  225.         When this is enabled, the backup will start by enumerating the size of

  226.         the entire database, and then go back and send the actual contents.

  227.         This may make the backup take slightly longer, and in particular it

  228.         will take longer before the first data is sent.

  229.        </p></dd><dt><span class="term"><code class="option">-S <em class="replaceable"><code>slotname</code></em></code><br /></span><span class="term"><code class="option">--slot=<em class="replaceable"><code>slotname</code></em></code></span></dt><dd><p>

  230.         This option can only be used together with <code class="literal">-X

  231.         stream</code>.  It causes the WAL streaming to use the specified

  232.         replication slot.  If the base backup is intended to be used as a

  233.         streaming replication standby using replication slots, it should then

  234.         use the same replication slot name

  235.         in <a class="xref" href="runtime-config-replication.html#GUC-PRIMARY-SLOT-NAME">primary_slot_name</a>.  That way, it is ensured that

  236.         the server does not remove any necessary WAL data in the time between

  237.         the end of the base backup and the start of streaming replication.

  238.        </p><p>

  239.         The specified replication slot has to exist unless the

  240.         option <code class="option">-C</code> is also used.

  241.        </p><p>

  242.         If this option is not specified and the server supports temporary

  243.         replication slots (version 10 and later), then a temporary replication

  244.         slot is automatically used for WAL streaming.

  245.        </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>

  246.         Enables verbose mode. Will output some extra steps during startup and

  247.         shutdown, as well as show the exact file name that is currently being

  248.         processed if progress reporting is also enabled.

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

  250.         This option prevents the creation of a temporary replication slot

  251.         during the backup even if it's supported by the server.

  252.        </p><p>

  253.         Temporary replication slots are created by default if no slot name

  254.         is given with the option <code class="option">-S</code> when using log streaming.

  255.        </p><p>

  256.         The main purpose of this option is to allow taking a base backup when

  257.         the server is out of free replication slots.  Using replication slots

  258.         is almost always preferred, because it prevents needed WAL from being

  259.         removed by the server during the backup.

  260.        </p></dd><dt><span class="term"><code class="option">--no-verify-checksums</code></span></dt><dd><p>

  261.         Disables verification of checksums, if they are enabled on the server

  262.         the base backup is taken from.

  263.        </p><p>

  264.         By default, checksums are verified and checksum failures will result

  265.         in a non-zero exit status. However, the base backup will not be

  266.         removed in such a case, as if the <code class="option">--no-clean</code> option

  267.         had been used.  Checksum verifications failures will also be reported

  268.         in the <a class="xref" href="monitoring-stats.html#PG-STAT-DATABASE-VIEW" title="Table 27.12. pg_stat_database View">pg_stat_database</a> view.

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

  270.    </p><p>

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

  272. 

  273.     </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>

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

  275.         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.

  276.        </p><p>

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

  278.         client applications, but because <span class="application">pg_basebackup</span>

  279.         doesn't connect to any particular database in the cluster, database

  280.         name in the connection string will be ignored.

  281.        </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>

  282.         Specifies the host name of the machine on which the server is

  283.         running.  If the value begins with a slash, it is used as the

  284.         directory for the Unix domain socket. The default is taken

  285.         from the <code class="envar">PGHOST</code> environment variable, if set,

  286.         else a Unix domain socket connection is attempted.

  287.        </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>

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

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

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

  291.         set, or a compiled-in default.

  292.        </p></dd><dt><span class="term"><code class="option">-s <em class="replaceable"><code>interval</code></em></code><br /></span><span class="term"><code class="option">--status-interval=<em class="replaceable"><code>interval</code></em></code></span></dt><dd><p>

  293.         Specifies the number of seconds between status packets sent back to the

  294.         server. This allows for easier monitoring of the progress from server.

  295.         A value of zero disables the periodic status updates completely,

  296.         although an update will still be sent when requested by the server, to

  297.         avoid timeout disconnect. The default value is 10 seconds.

  298.        </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>

  299.         User name to connect as.

  300.        </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>

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

  302.         password authentication and a password is not available by

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

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

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

  306.         password.

  307.        </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>

  308.         Force <span class="application">pg_basebackup</span> to prompt for a

  309.         password before connecting to a database.

  310.        </p><p>

  311.         This option is never essential, since

  312.         <span class="application">pg_basebackup</span> will automatically prompt

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

  314.         However, <span class="application">pg_basebackup</span> will waste a

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

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

  317.         connection attempt.

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

  319.    </p><p>

  320.     Other options are also available:

  321. 

  322.     </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="option">-V</code><br /></span><span class="term"><code class="option">--version</code></span></dt><dd><p>

  323.        Print the <span class="application">pg_basebackup</span> version and exit.

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

  325.        Show help about <span class="application">pg_basebackup</span> command line

  326.        arguments, and exit.

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

  328.    </p></div><div class="refsect1" id="id-1.9.4.9.7"><h2>Environment</h2><p>

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

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

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

  332.   </p><p>

  333.    The environment variable <code class="envar">PG_COLOR</code> specifies whether to use

  334.    color in diagnostic messages. Possible values are

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

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

  337.   </p></div><div class="refsect1" id="id-1.9.4.9.8"><h2>Notes</h2><p>

  338.    At the beginning of the backup, a checkpoint needs to be written on the

  339.    server the backup is taken from.  Especially if the option

  340.    <code class="literal">--checkpoint=fast</code> is not used, this can take some time

  341.    during which <span class="application">pg_basebackup</span> will be appear

  342.    to be idle.

  343.   </p><p>

  344.    The backup will include all files in the data directory and tablespaces,

  345.    including the configuration files and any additional files placed in the

  346.    directory by third parties, except certain temporary files managed by

  347.    PostgreSQL.  But only regular files and directories are copied, except that

  348.    symbolic links used for tablespaces are preserved.  Symbolic links pointing

  349.    to certain directories known to PostgreSQL are copied as empty directories.

  350.    Other symbolic links and special device files are skipped.

  351.    See <a class="xref" href="protocol-replication.html" title="52.4. Streaming Replication Protocol">Section 52.4</a> for the precise details.

  352.   </p><p>

  353.    Tablespaces will in plain format by default be backed up to the same path

  354.    they have on the server, unless the

  355.    option <code class="literal">--tablespace-mapping</code> is used.  Without

  356.    this option, running a plain format base backup on the same host as the

  357.    server will not work if tablespaces are in use, because the backup would

  358.    have to be written to the same directory locations as the original

  359.    tablespaces.

  360.   </p><p>

  361.    When tar format mode is used, it is the user's responsibility to unpack each

  362.    tar file before starting the PostgreSQL server. If there are additional tablespaces, the

  363.    tar files for them need to be unpacked in the correct locations. In this

  364.    case the symbolic links for those tablespaces will be created by the server

  365.    according to the contents of the <code class="filename">tablespace_map</code> file that is

  366.    included in the <code class="filename">base.tar</code> file.

  367.   </p><p>

  368.    <span class="application">pg_basebackup</span> works with servers of the same

  369.    or an older major version, down to 9.1. However, WAL streaming mode (<code class="literal">-X

  370.    stream</code>) only works with server version 9.3 and later, and tar format mode

  371.    (<code class="literal">--format=tar</code>) of the current version only works with server version 9.5

  372.    or later.

  373.   </p><p>

  374.    <span class="application">pg_basebackup</span> will preserve group permissions in

  375.    both the <code class="literal">plain</code> and <code class="literal">tar</code> formats if group

  376.    permissions are enabled on the source cluster.

  377.   </p></div><div class="refsect1" id="id-1.9.4.9.9"><h2>Examples</h2><p>

  378.    To create a base backup of the server at <code class="literal">mydbserver</code>

  379.    and store it in the local directory

  380.    <code class="filename">/usr/local/pgsql/data</code>:

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

  382. <code class="prompt">$</code> <strong class="userinput"><code>pg_basebackup -h mydbserver -D /usr/local/pgsql/data</code></strong>

  383. </pre><p>

  384.   </p><p>

  385.    To create a backup of the local server with one compressed

  386.    tar file for each tablespace, and store it in the directory

  387.    <code class="filename">backup</code>, showing a progress report while running:

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

  389. <code class="prompt">$</code> <strong class="userinput"><code>pg_basebackup -D backup -Ft -z -P</code></strong>

  390. </pre><p>

  391.   </p><p>

  392.    To create a backup of a single-tablespace local database and compress

  393.    this with <span class="productname">bzip2</span>:

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

  395. <code class="prompt">$</code> <strong class="userinput"><code>pg_basebackup -D - -Ft -X fetch | bzip2 &gt; backup.tar.bz2</code></strong>

  396. </pre><p>

  397.    (This command will fail if there are multiple tablespaces in the

  398.    database.)

  399.   </p><p>

  400.    To create a backup of a local database where the tablespace in

  401.    <code class="filename">/opt/ts</code> is relocated

  402.    to <code class="filename">./backup/ts</code>:

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

  404. <code class="prompt">$</code> <strong class="userinput"><code>pg_basebackup -D backup/data -T /opt/ts=$(pwd)/backup/ts</code></strong>

  405. </pre><p>

  406.   </p></div><div class="refsect1" id="id-1.9.4.9.10"><h2>See Also</h2><span class="simplelist"><a class="xref" href="app-pgdump.html" title="pg_dump"><span class="refentrytitle">pg_dump</span></a></span></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="app-ecpg.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="pgbench.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top"><span class="application">ecpg</span> </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">pgbench</span></td></tr></table></div></body></html>
上海开阖软件有限公司 沪ICP备12045867号-1