<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!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>25.2. File System Level Backup</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 Vsnapshot" /><link rel="prev" href="backup-dump.html" title="25.1. SQL Dump" /><link rel="next" href="continuous-archiving.html" title="25.3. Continuous Archiving and Point-in-Time Recovery (PITR)" /></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">25.2. File System Level Backup</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="backup-dump.html" title="25.1. SQL Dump">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="backup.html" title="Chapter 25. Backup and Restore">Up</a></td><th width="60%" align="center">Chapter 25. Backup and Restore</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 11.5 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="continuous-archiving.html" title="25.3. Continuous Archiving and Point-in-Time Recovery (PITR)">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="BACKUP-FILE"><div class="titlepage"><div><div><h2 class="title" style="clear: both">25.2. File System Level Backup</h2></div></div></div><p>
An alternative backup strategy is to directly copy the files that
<span class="productname">PostgreSQL</span> uses to store the data in the database;
<a class="xref" href="creating-cluster.html" title="18.2. Creating a Database Cluster">Section 18.2</a> explains where these files
are located. You can use whatever method you prefer
for doing file system backups; for example:
</p><pre class="programlisting">
tar -cf backup.tar /usr/local/pgsql/data
</pre><p>
</p><p>
There are two restrictions, however, which make this method
impractical, or at least inferior to the <span class="application">pg_dump</span>
method:
</p><div class="orderedlist"><ol class="orderedlist" type="1"><li class="listitem"><p>
The database server <span class="emphasis"><em>must</em></span> be shut down in order to
get a usable backup. Half-way measures such as disallowing all
connections will <span class="emphasis"><em>not</em></span> work
(in part because <code class="command">tar</code> and similar tools do not take
an atomic snapshot of the state of the file system,
but also because of internal buffering within the server).
Information about stopping the server can be found in
<a class="xref" href="server-shutdown.html" title="18.5. Shutting Down the Server">Section 18.5</a>. Needless to say, you
also need to shut down the server before restoring the data.
</p></li><li class="listitem"><p>
If you have dug into the details of the file system layout of the
database, you might be tempted to try to back up or restore only certain
individual tables or databases from their respective files or
directories. This will <span class="emphasis"><em>not</em></span> work because the
information contained in these files is not usable without
the commit log files,
<code class="filename">pg_xact/*</code>, which contain the commit status of
all transactions. A table file is only usable with this
information. Of course it is also impossible to restore only a
table and the associated <code class="filename">pg_xact</code> data
because that would render all other tables in the database
cluster useless. So file system backups only work for complete
backup and restoration of an entire database cluster.
</p></li></ol></div><p>
</p><p>
An alternative file-system backup approach is to make a
<span class="quote">“<span class="quote">consistent snapshot</span>”</span> of the data directory, if the
file system supports that functionality (and you are willing to
trust that it is implemented correctly). The typical procedure is
to make a <span class="quote">“<span class="quote">frozen snapshot</span>”</span> of the volume containing the
database, then copy the whole data directory (not just parts, see
above) from the snapshot to a backup device, then release the frozen
snapshot. This will work even while the database server is running.
However, a backup created in this way saves
the database files in a state as if the database server was not
properly shut down; therefore, when you start the database server
on the backed-up data, it will think the previous server instance
crashed and will replay the WAL log. This is not a problem; just
be aware of it (and be sure to include the WAL files in your backup).
You can perform a <code class="command">CHECKPOINT</code> before taking the
snapshot to reduce recovery time.
</p><p>
If your database is spread across multiple file systems, there might not
be any way to obtain exactly-simultaneous frozen snapshots of all
the volumes. For example, if your data files and WAL log are on different
disks, or if tablespaces are on different file systems, it might
not be possible to use snapshot backup because the snapshots
<span class="emphasis"><em>must</em></span> be simultaneous.
Read your file system documentation very carefully before trusting
the consistent-snapshot technique in such situations.
</p><p>
If simultaneous snapshots are not possible, one option is to shut down
the database server long enough to establish all the frozen snapshots.
Another option is to perform a continuous archiving base backup (<a class="xref" href="continuous-archiving.html#BACKUP-BASE-BACKUP" title="25.3.2. Making a Base Backup">Section 25.3.2</a>) because such backups are immune to file
system changes during the backup. This requires enabling continuous
archiving just during the backup process; restore is done using
continuous archive recovery (<a class="xref" href="continuous-archiving.html#BACKUP-PITR-RECOVERY" title="25.3.4. Recovering Using a Continuous Archive Backup">Section 25.3.4</a>).
</p><p>
Another option is to use <span class="application">rsync</span> to perform a file
system backup. This is done by first running <span class="application">rsync</span>
while the database server is running, then shutting down the database
server long enough to do an <code class="command">rsync --checksum</code>.
(<code class="option">--checksum</code> is necessary because <code class="command">rsync</code> only
has file modification-time granularity of one second.) The
second <span class="application">rsync</span> will be quicker than the first,
because it has relatively little data to transfer, and the end result
will be consistent because the server was down. This method
allows a file system backup to be performed with minimal downtime.
</p><p>
Note that a file system backup will typically be larger
than an SQL dump. (<span class="application">pg_dump</span> does not need to dump
the contents of indexes for example, just the commands to recreate
them.) However, taking a file system backup might be faster.
</p></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="backup-dump.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="backup.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="continuous-archiving.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">25.1. <acronym class="acronym">SQL</acronym> Dump </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> 25.3. Continuous Archiving and Point-in-Time Recovery (PITR)</td></tr></table></div></body></html>
