</listitem>

</itemizedlist>

<para><emphasis role="strong">See Also</emphasis></para>

<para><xref linkend="pg-probackup-creating-backup"></para>

<para><link linkend="pg-probackup-creating-backup">Creating Backups</link></para>

<refsect2 id="pg-probackup-limitations">

<title>Limitations</title>

</listitem>

<listitem>

<para>

<link linkend="configuring-the-database-cluster-for-backups">Configure the database clusterfor backups</link>.

<link linkend="configuring-database-cluster">Configure the database clusterto enable <application>pg_probackup</application> backups</link>.

</para>

</listitem>

</orderedlist>

For details, see <xref linkend="configuring-pg-probackup">.

</para>

</refsect2>

<refsect2 id="configuring-the-database-cluster-for-backups">

<title>Configuring the Database Cluster for Backups</title>

<refsect2 id="configuring-database-cluster">

<title>Configuring the Database Cluster</title>

<para>Depending on whether you are going to use autonomous or archive backup

strategies, <productname>&productname;</productname> cluster configuration

will differ. To backup the database cluster from a standby

<orderedlist numeration="arabic">

<listitem>

<para>

Create the <literal>backup</literal> role withthe

Create the <literal>backup</literal> role with

<link linkend="role-attributes"><literal>LOGIN</literal> and <literal>REPLICATION</literal> privileges</link>.

</para>

<programlisting>

<listitem>

<para>

Set <varname>archive_command</varname> to

<literal>test ! -fbackup_dir/wal/%f &amp;&amp; cp %pbackup_dir/wal/%f</literal>

where <replaceable>backup_dir</replaceable> is the initialized

<literal>test ! -fbackupdir/wal/%f &amp;&amp; cp %pbackupdir/wal/%f</literal>

where <replaceable>backupdir</replaceable> is the initialized

backup catalog.

</para>

</listitem>

<title>Creating a Backup</title>

<para>To create a backup, run the following command:

<programlisting>

pg_probackup backup -b <replaceable>backup_mode</replaceable>

pg_probackup backup -B <replaceable>backupdir</replaceable> -b <replaceable>backup_mode</replaceable>

</programlisting>

where <replaceable>backup_mode</replaceable> can take one of the following values:</para>

<itemizedlist spacing="compact">

To make a backup autonomous, add the <option>--stream</option> option to the above command. For example, to create a full autonomous backup, run:

</para>

<programlisting>

pg_probackup backup -b FULL --stream

pg_probackup backup -B <replaceable>backupdir</replaceable> -b FULL --stream

</programlisting>

<para>

<replaceable>xid</replaceable> transaction ID, run this command:

</para>

<programlisting>

pg_probackup validate --xid=<replaceable>xid</replaceable>

pg_probackup validate -B <replaceable>backupdir</replaceable> --xid=<replaceable>xid</replaceable>

</programlisting>

<para>

If validation completes successfully,

</listitem>

</orderedlist>

<programlisting>

pg_probackup restore -D <replaceable>datadir</replaceable> -i <replaceable>backup_id</replaceable>

pg_probackup restore -B <replaceable>backupdir</replaceable> -D <replaceable>datadir</replaceable> -i <replaceable>backup_id</replaceable>

</programlisting>

<para>

where:

format. For example:

</para>

<programlisting>

pg_probackup restore --time='2017-05-18 14:18:11'

pg_probackup restore -B <replaceable>backupdir</replaceable> --time='2017-05-18 14:18:11'

</programlisting>

<para>

To restore the cluster state up to a specific transaction ID, use the <option>xid</option>:

</para>

<programlisting>

pg_probackup restore --xid=687

pg_probackup restore -B <replaceable>backupdir</replaceable> --xid=687

</programlisting>

<para>

If the cluster to restore contains tablespaces,

tablespaces. For example:

</para>

<programlisting>

pg_probackup restore -D <replaceable>datadir</replaceable> -j 4 -i <replaceable>backup_id</replaceable> -T <replaceable>/tmp/tablespace1</replaceable>=<replaceable>/tmp/tablespace1_new</replaceable> -T <replaceable>/tmp/tablespace2</replaceable>=<replaceable>/tmp/tablespace2_new</replaceable>

pg_probackup restore -B <replaceable>backupdir</replaceable> -D <replaceable>datadir</replaceable> -j 4 -i <replaceable>backup_id</replaceable> -T <replaceable>tablespace1_dir</replaceable>=<replaceable>tablespace1_newdir</replaceable> -T <replaceable>tablespace2_dir</replaceable>=<replaceable>tablespace2_newdir</replaceable>

</programlisting>

<para>

Once the <literal>restore</literal> command is complete, start the

Parallel execution is controlled by the <option>-j</option>/<option>--threads</option>

command line option. For example, to create a backup using four parallel threads, run:

<programlisting>

pg_probackup backup -b full -j 4

pg_probackup backup -B <replaceable>backupdir</replaceable> -b FULL -j 4

</programlisting>

</para>

<refsect1 id="managing-backup-catalog">

<title>Managing the Backup Catalog</title>

<para>

With <application>pg_probackup</application>, you can manage and configure

With <application>pg_probackup</application>, you can manage

backups from the command line:

</para>

<itemizedlist>

<listitem>

<para><xref linkend="viewing-backup-info">

<para><link linkend="viewing-backup-info">View available backups</link>

</para>

</listitem>

<listitem>

<para><xref linkend="pg-probackup-deleting-backups">

<para><link linkend="pg-probackup-deleting-backups">Delete backups</link>

</para>

</listitem>

</itemizedlist>