<title>Health Check</title>
<para>
- health_check_period, health_check_timeout, health_check_user,
- health_check_password, health_check_database, health_check_max_retries,
- health_check_retry_delay, connect_timeout
+ <productname>Pgpool-II</productname> periodically connects to the configured
+ PostgreSQL backends to detect any error on the servers or networks.
+ This error check procedure is called "health check".
+ If an error is detected, <productname>Pgpool-II</productname> performs failover
+ or degeneration depending on the configurations.
+ <caution>
+ <para>
+ Health check requires one extra connection to each backend,
+ so max_connections in the <filename>postgresql.conf</filename>
+ needs to be adjusted accordingly.
+ </para>
+ </caution>
</para>
+
+ <variablelist>
+
+ <varlistentry id="guc-health-check-timeout" xreflabel="health_check_timeout">
+ <term><varname>health_check_timeout</varname> (<type>integer</type>)
+ <indexterm>
+ <primary><varname>health_check_timeout</varname> configuration parameter</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ Specifies the timeout in seconds to give up connecting to the backend
+ PostgreSQL if the TCP connect does not succeed within this time.
+ </para>
+ <para>
+ This parameter serves to prevent the health check from waiting for a
+ long time in a case such as unplugged network cable.
+ Default value is 20. Setting it to 0, disables the timeout (waits until TCP/IP timeout).
+ </para>
+ <para>
+ This parameter can be changed by reloading pgpool configuration.
+ </para>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry id="guc-health-check-period" xreflabel="health_check_period">
+ <term><varname>health_check_period</varname> (<type>integer</type>)
+ <indexterm>
+ <primary><varname>health_check_period</varname> configuration parameter</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ Specifies the interval between the health checks in seconds.
+ Default is 0, which means health check is disabled.
+ </para>
+ <para>
+ This parameter can be changed by reloading pgpool configuration.
+ </para>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry id="guc-health_check_user" xreflabel="health_check_user">
+ <term><varname>health_check_user</varname> (<type>string</type>)
+ <indexterm>
+ <primary><varname>health_check_user</varname> configuration parameter</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ Specifies the PostgreSQL user name to perform health check.
+ The same user must exist in all the PostgreSQL backends.
+ Otherwise, health check causes an error.
+ </para>
+ <para>
+ This parameter can be changed by reloading pgpool configuration.
+ </para>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry id="guc-health-check-password" xreflabel="health_check_password">
+ <term><varname>health_check_password</varname> (<type>string</type>)
+ <indexterm>
+ <primary><varname>health_check_password</varname> configuration parameter</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ Specifies the password for the PostgreSQL user name configured in
+ <xref linkend="guc-health_check_user"> to perform health check.
+ The user and password must be same in all the PostgreSQL backends.
+ Otherwise, health check causes an error.
+ </para>
+ <para>
+ This parameter can be changed by reloading pgpool configuration.
+ </para>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry id="guc-health-check-database" xreflabel="health_check_database">
+ <term><varname>health_check_database</varname> (<type>string</type>)
+ <indexterm>
+ <primary><varname>health_check_database</varname> configuration parameter</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ Specifies the PostgreSQL database name to perform health check.
+ The default is <literal>''</literal>(empty), which tries <literal>"postgres"</literal>
+ database first, then <literal>"template1"</literal> database until it succeeds
+ </para>
+ <para>
+ <varname>health_check_database</varname> was introduced in
+ <productname>Pgpool-II</productname> <emphasis>V3.5</emphasis>.
+ </para>
+ <para>
+ This parameter can be changed by reloading pgpool configuration.
+ </para>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry id="guc-health-check-max-retries" xreflabel="health_check_max_retries">
+ <term><varname>health_check_max_retries</varname> (<type>integer</type>)
+ <indexterm>
+ <primary><varname>health_check_max_retries</varname> configuration parameter</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ Specifies the maximum number of retries to do with a failed health check
+ before giving up and initiating failover.
+ <tip>
+ <para>
+ This setting can be useful in spotty networks, when it is expected that
+ health checks will fail occasionally even when the master is fine.
+ It is advised that you disable fail_over_on_backend_error if you want to enable health_check_max_retries.
+ </para>
+ </tip>
+ <tip>
+ <para>
+ It is advised that fail_over_on_backend_error must be disabled,
+ if you want to enable health_check_max_retries.
+ </para>
+ </tip>
+ Default is 0, which means do not retry.
+ </para>
+ <para>
+ This parameter can be changed by reloading pgpool configuration.
+ </para>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry id="guc-health-check-retry-delay" xreflabel="health_check_retry_delay">
+ <term><varname>health_check_retry_delay</varname> (<type>integer</type>)
+ <indexterm>
+ <primary><varname>health_check_retry_delay</varname> configuration parameter</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ Specifies the amount of time (in seconds) to sleep between failed
+ health check retries (not used unless <xref linkend="guc-health-check-max-retries"> is > 0).
+ If 0, then retries are immediate (no delay).
+ </para>
+ <para>
+ This parameter can be changed by reloading pgpool configuration.
+ </para>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry id="guc-connect-timeout" xreflabel="connect_timeout">
+ <term><varname>connect_timeout</varname> (<type>integer</type>)
+ <indexterm>
+ <primary><varname>connect_timeout</varname> configuration parameter</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ Specifies the amount of time in milliseconds before giving up connecting
+ to backend using connect() system call.
+ Default is 10000 ms (10 second). The flaky network user may want to increase the value.
+ 0 means no timeout.
+ <note>
+ <para>
+ <varname>connect_timeout</varname> value is not only used for a health check,
+ but also for creating ordinary connection pools.
+ </para>
+ </note>
+ </para>
+ <para>
+ This parameter can be changed by reloading pgpool configuration.
+ </para>
+ </listitem>
+ </varlistentry>
+
+ </variablelist>
</sect1>
projects / pgpool2.git / commitdiff