<sect3 id="multimaster-guc-variables">

<title>GUC Variables</title>

<variablelist>

<varlistentry><term><varname>multimaster.node_id</varname>

<varlistentry id="multimaster-node-id"><term><varname>multimaster.node_id</varname>

<indexterm><primary><varname>multimaster.node_id</varname></primary></indexterm></term><listitem><para>

Node ID &mdash; a unique natural

number identifying the node of a multi-master cluster. You must

schedule jobs, in the comma-separated format:

<programlisting>

<varname>schedule.database</varname> = '<replaceable>database1</>,<replaceable>database2</>'

</programlisting></para>

</programlisting>

You can omit this step when using <filename>pgpro_scheduler</filename>

on a cluster configured with <filename>multimaster</filename>.</para>

</listitem>

<listitem>

<para>

</term>

<listitem>

<para>

Specifies whetherthe scheduler is enabled on your system.</para>

Specifies whether<filename>pgpro_scheduler</filename> is enabled on your system.</para>

<para>Default: <literal>false</literal>.

</para>

</listitem>

</term>

<listitem>

<para>

Specifies the list of databases for which the scheduler is enabled. Database names must be separated by commas.</para><para>Default: empty string.

Specifies the databases for which <filename>pgpro_scheduler</filename> is enabled.

Database names must be separated by commas.

</para>

<para>

If <filename>pgpro_scheduler</filename> is running on a cluster configured with

<filename>multimaster</filename>, the database name is derived from the

<xref linkend="multimaster-conn-strings"> variable.

</para>

<para>

Default: empty string.

</para>

</listitem>

</varlistentry>

</term>

<listitem>

<para>

Specifies the name of a schema where the scheduler

stores its tables and functions. Normally, you should

not change this variable. This

variable may be useful if you run <filename>pgpro_scheduler</filename>

on a multi-master cluster.</para>

Deprecated. Specifies the name of a schema where the scheduler

stores its tables and functions.

If you need to change the default schema, use <xref linkend="sql-alterextension">.

</para>

<para>Default: <literal>schedule</literal>.

</para>

</listitem>

</term>

<listitem>

<para>

Specifies the node name of your <productname>&productname;</productname> instance. Do

not change or use this variable if you run a single server configuration. You may need this

variable if you run <filename>pgpro_scheduler</filename>

on a multi-master cluster.</para>

Specifies the node name of your <productname>&productname;</productname> instance.

Do not change or use this variable if you run a single-server cluster configuration.

</para>

<para>

On a cluster configured with <filename>multimaster</filename>, the node name is

derived from the <xref linkend="multimaster-node-id"> variable. For example, if

<literal>multimaster.node_id = 3</literal>, the <varname>schedule.nodename</>

variable is set to <literal>mtm-node-3</>. However, if you explicitly set

the <varname>schedule.nodename</> variable by editing the <filename>postgresql.conf</filename>

file or running the <command>ALTER</command> command, <filename>pgpro_scheduler</filename>

will use the provided value, ignoring the <varname>multimaster.node_id</varname> variable.

</para>

<para>Default: <literal>master</literal>.

</para>

</listitem>

<para>The number of records deleted.</para></listitem></itemizedlist>

</listitem>

</varlistentry>

<varlistentry>

<term>

<function>schedule.nodename()

</function>

<indexterm>

<primary><function>schedule.nodename()</></primary>

</indexterm>

</term>

<listitem>

<para>

Returns the current node name.

</para>

</listitem>

</varlistentry>

</variablelist>

</sect4>

<sect4 id="pgpro-scheduler-one-time-job-functions"><title>Functions for Managing One-Time Jobs</title>