<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<HTML
><HEAD
><TITLE
>Run-time Configuration</TITLE
><META
NAME="GENERATOR"
CONTENT="Modular DocBook HTML Stylesheet Version 1.79"><LINK
REV="MADE"
HREF="mailto:pgsql-docs@postgresql.org"><LINK
REL="HOME"
TITLE="PostgreSQL 8.0.11 Documentation"
HREF="index.html"><LINK
REL="UP"
TITLE="Server Run-time Environment"
HREF="runtime.html"><LINK
REL="PREVIOUS"
TITLE="Starting the Database Server"
HREF="postmaster-start.html"><LINK
REL="NEXT"
TITLE="Managing Kernel Resources"
HREF="kernel-resources.html"><LINK
REL="STYLESHEET"
TYPE="text/css"
HREF="stylesheet.css"><META
HTTP-EQUIV="Content-Type"
CONTENT="text/html; charset=ISO-8859-1"><META
NAME="creation"
CONTENT="2007-02-02T03:57:22"></HEAD
><BODY
CLASS="SECT1"
><DIV
CLASS="NAVHEADER"
><TABLE
SUMMARY="Header navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TH
COLSPAN="5"
ALIGN="center"
VALIGN="bottom"
>PostgreSQL 8.0.11 Documentation</TH
></TR
><TR
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
HREF="postmaster-start.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
HREF="runtime.html"
>Fast Backward</A
></TD
><TD
WIDTH="60%"
ALIGN="center"
VALIGN="bottom"
>Chapter 16. Server Run-time Environment</TD
><TD
WIDTH="10%"
ALIGN="right"
VALIGN="top"
><A
HREF="runtime.html"
>Fast Forward</A
></TD
><TD
WIDTH="10%"
ALIGN="right"
VALIGN="top"
><A
HREF="kernel-resources.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
></TABLE
><HR
ALIGN="LEFT"
WIDTH="100%"></DIV
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="RUNTIME-CONFIG"
>16.4. Run-time Configuration</A
></H1
><A
NAME="AEN16617"
></A
><P
> There are a lot of configuration parameters that affect the
behavior of the database system. In this subsection, we describe
how to set configuration parameters; the following subsections
discuss each parameter in detail.
</P
><P
> All parameter names are case-insensitive. Every parameter takes a
value of one of four types: boolean, integer, floating point,
or string. Boolean values may be written as <TT
CLASS="LITERAL"
>ON</TT
>,
<TT
CLASS="LITERAL"
>OFF</TT
>, <TT
CLASS="LITERAL"
>TRUE</TT
>,
<TT
CLASS="LITERAL"
>FALSE</TT
>, <TT
CLASS="LITERAL"
>YES</TT
>,
<TT
CLASS="LITERAL"
>NO</TT
>, <TT
CLASS="LITERAL"
>1</TT
>, <TT
CLASS="LITERAL"
>0</TT
>
(all case-insensitive) or any unambiguous prefix of these.
</P
><P
> One way to set these parameters is to edit the file
<TT
CLASS="FILENAME"
>postgresql.conf</TT
><A
NAME="AEN16632"
></A
>,
which is normally kept in the data directory. (<TT
CLASS="COMMAND"
>initdb</TT
>
installs a default copy there.) An example of what this file might look
like is:
</P><PRE
CLASS="PROGRAMLISTING"
># This is a comment
log_connections = yes
log_destination = 'syslog'
search_path = '$user, public'</PRE
><P>
One parameter is specified per line. The equal sign between name and
value is optional. Whitespace is insignificant and blank lines are
ignored. Hash marks (<TT
CLASS="LITERAL"
>#</TT
>) introduce comments
anywhere. Parameter values that are not simple identifiers or
numbers must be single-quoted.
</P
><P
> <A
NAME="AEN16638"
></A
>
The configuration file is reread whenever the
<TT
CLASS="COMMAND"
>postmaster</TT
> process receives a
<SPAN
CLASS="SYSTEMITEM"
>SIGHUP</SPAN
> signal (which is most easily sent by means
of <TT
CLASS="LITERAL"
>pg_ctl reload</TT
>). The <TT
CLASS="COMMAND"
>postmaster</TT
>
also propagates this signal to all currently running server
processes so that existing sessions also get the new
value. Alternatively, you can send the signal to a single server
process directly. Some parameters can only be set at server start;
any changes to their entries in the configuration file will be ignored
until the server is restarted.
</P
><P
> A second way to set these configuration parameters is to give them
as a command line option to the <TT
CLASS="COMMAND"
>postmaster</TT
>, such as:
</P><PRE
CLASS="PROGRAMLISTING"
>postmaster -c log_connections=yes -c log_destination='syslog'</PRE
><P>
Command-line options override any conflicting settings in
<TT
CLASS="FILENAME"
>postgresql.conf</TT
>. Note that this means you won't
be able to change the value on-the-fly by editing
<TT
CLASS="FILENAME"
>postgresql.conf</TT
>, so while the command-line
method may be convenient, it can cost you flexibility later.
</P
><P
> Occasionally it is useful to give a command line option to
one particular session only. The environment variable
<TT
CLASS="ENVAR"
>PGOPTIONS</TT
> can be used for this purpose on the
client side:
</P><PRE
CLASS="PROGRAMLISTING"
>env PGOPTIONS='-c geqo=off' psql</PRE
><P>
(This works for any <SPAN
CLASS="APPLICATION"
>libpq</SPAN
>-based client application, not
just <SPAN
CLASS="APPLICATION"
>psql</SPAN
>.) Note that this won't work for
parameters that are fixed when the server is started or that must be
specified in <TT
CLASS="FILENAME"
>postgresql.conf</TT
>.
</P
><P
> Furthermore, it is possible to assign a set of option settings to
a user or a database. Whenever a session is started, the default
settings for the user and database involved are loaded. The
commands <A
HREF="sql-alteruser.html"
><I
>ALTER USER</I
></A
>
and <A
HREF="sql-alterdatabase.html"
><I
>ALTER DATABASE</I
></A
>,
respectively, are used to configure these settings. Per-database
settings override anything received from the
<TT
CLASS="COMMAND"
>postmaster</TT
> command-line or the configuration
file, and in turn are overridden by per-user settings; both are
overridden by per-session options.
</P
><P
> Some parameters can be changed in individual <ACRONYM
CLASS="ACRONYM"
>SQL</ACRONYM
>
sessions with the <A
HREF="sql-set.html"
><I
>SET</I
></A
>
command, for example:
</P><PRE
CLASS="SCREEN"
>SET ENABLE_SEQSCAN TO OFF;</PRE
><P>
If <TT
CLASS="COMMAND"
>SET</TT
> is allowed, it overrides all other sources of
values for the parameter. Some parameters cannot be changed via
<TT
CLASS="COMMAND"
>SET</TT
>: for example, if they control behavior that
cannot reasonably be changed without restarting
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>. Also, some parameters can
be modified via <TT
CLASS="COMMAND"
>SET</TT
> or <TT
CLASS="COMMAND"
>ALTER</TT
> by superusers,
but not by ordinary users.
</P
><P
> The <A
HREF="sql-show.html"
><I
>SHOW</I
></A
>
command allows inspection of the current values of all parameters.
</P
><P
> The virtual table <TT
CLASS="STRUCTNAME"
>pg_settings</TT
>
(described in <A
HREF="view-pg-settings.html"
>Section 41.35</A
>) also allows
displaying and updating session run-time parameters. It is equivalent
to <TT
CLASS="COMMAND"
>SHOW</TT
> and <TT
CLASS="COMMAND"
>SET</TT
>, but can be more convenient
to use because it can be joined with other tables, or selected from using
any desired selection condition.
</P
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-FILE-LOCATIONS"
>16.4.1. File Locations</A
></H2
><P
> In addition to the <TT
CLASS="FILENAME"
>postgresql.conf</TT
> file
already mentioned, <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> uses
two other manually-edited configuration files, which control
client authentication (their use is discussed in <A
HREF="client-authentication.html"
>Chapter 19</A
>).
By default, all three configuration files are stored
in the database cluster's data directory. The options described
in this subsection allow the configuration files to be placed elsewhere.
(Doing so can ease administration. In particular it is often
easier to ensure that the configuration files are properly backed-up
when they are kept separate.)
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-DATA-DIRECTORY"
></A
><TT
CLASS="VARNAME"
>data_directory</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Specifies the directory to use for data storage.
This option can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-CONFIG-FILE"
></A
><TT
CLASS="VARNAME"
>config_file</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Specifies the main server configuration file
(customarily called <TT
CLASS="FILENAME"
>postgresql.conf</TT
>).
This option can only be set on the postmaster command line.
</P
></DD
><DT
><A
NAME="GUC-HBA-FILE"
></A
><TT
CLASS="VARNAME"
>hba_file</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Specifies the configuration file for host-based authentication
(customarily called <TT
CLASS="FILENAME"
>pg_hba.conf</TT
>).
This option can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-IDENT-FILE"
></A
><TT
CLASS="VARNAME"
>ident_file</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Specifies the configuration file for
<SPAN
CLASS="APPLICATION"
>ident</SPAN
> authentication
(customarily called <TT
CLASS="FILENAME"
>pg_ident.conf</TT
>).
This option can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-EXTERNAL-PID-FILE"
></A
><TT
CLASS="VARNAME"
>external_pid_file</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Specifies the name of an additional process-id (PID) file that the
<SPAN
CLASS="APPLICATION"
>postmaster</SPAN
> should create for use by server
administration programs.
This option can only be set at server start.
</P
></DD
></DL
></DIV
><P
> In a default installation, none of the above options are set explicitly.
Instead, the
data directory is specified by the <TT
CLASS="OPTION"
>-D</TT
> command-line
option or the <TT
CLASS="ENVAR"
>PGDATA</TT
> environment variable, and the
configuration files are all found within the data directory.
</P
><P
> If you wish to keep the configuration files elsewhere than the
data directory, the postmaster's <TT
CLASS="OPTION"
>-D</TT
>
command-line option or <TT
CLASS="ENVAR"
>PGDATA</TT
> environment variable
must point to the directory containing the configuration files,
and the <TT
CLASS="VARNAME"
>data_directory</TT
> option must be set in
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> (or on the command line) to show
where the data directory is actually located. Notice that
<TT
CLASS="VARNAME"
>data_directory</TT
> overrides <TT
CLASS="OPTION"
>-D</TT
> and
<TT
CLASS="ENVAR"
>PGDATA</TT
> for the location
of the data directory, but not for the location of the configuration
files.
</P
><P
> If you wish, you can specify the configuration file names and locations
individually using the options <TT
CLASS="VARNAME"
>config_file</TT
>,
<TT
CLASS="VARNAME"
>hba_file</TT
> and/or <TT
CLASS="VARNAME"
>ident_file</TT
>.
<TT
CLASS="VARNAME"
>config_file</TT
> can only be specified on the
<TT
CLASS="COMMAND"
>postmaster</TT
> command line, but the others can be
set within the main configuration file. If all three options plus
<TT
CLASS="VARNAME"
>data_directory</TT
> are explicitly set, then it is not necessary
to specify <TT
CLASS="OPTION"
>-D</TT
> or <TT
CLASS="ENVAR"
>PGDATA</TT
>.
</P
><P
> When setting any of these options, a relative path will be interpreted
with respect to the directory in which the <TT
CLASS="COMMAND"
>postmaster</TT
>
is started.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-CONNECTION"
>16.4.2. Connections and Authentication</A
></H2
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-CONNECTION-SETTINGS"
>16.4.2.1. Connection Settings</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-LISTEN-ADDRESSES"
></A
><TT
CLASS="VARNAME"
>listen_addresses</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Specifies the TCP/IP address(es) on which the server is
to listen for connections from client applications.
The value takes the form of a comma-separated list of host names
and/or numeric IP addresses. The special entry <TT
CLASS="LITERAL"
>*</TT
>
corresponds to all available IP interfaces.
If the list is empty, the server does not listen on any IP interface
at all, in which case only Unix-domain sockets can be used to connect
to it.
The default value is <SPAN
CLASS="SYSTEMITEM"
>localhost</SPAN
>,
which allows only local <SPAN
CLASS="QUOTE"
>"loopback"</SPAN
> connections to be made.
This parameter can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-PORT"
></A
><TT
CLASS="VARNAME"
>port</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> The TCP port the server listens on; 5432 by default. Note that the
same port number is used for all IP addresses the server listens on.
This parameter can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-MAX-CONNECTIONS"
></A
><TT
CLASS="VARNAME"
>max_connections</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Determines the maximum number of concurrent connections to the
database server. The default is typically 100, but may be less
if your kernel settings will not support it (as determined
during <SPAN
CLASS="APPLICATION"
>initdb</SPAN
>). This parameter can only be
set at server start.
</P
><P
> Increasing this parameter may cause <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>
to request more <SPAN
CLASS="SYSTEMITEM"
>System V</SPAN
> shared
memory or semaphores than your operating system's default configuration
allows. See <A
HREF="kernel-resources.html#SYSVIPC"
>Section 16.5.1</A
> for information on how to
adjust those parameters, if necessary.
</P
></DD
><DT
><A
NAME="GUC-SUPERUSER-RESERVED-CONNECTIONS"
></A
><TT
CLASS="VARNAME"
>superuser_reserved_connections</TT
>
(<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Determines the number of connection <SPAN
CLASS="QUOTE"
>"slots"</SPAN
> that
are reserved for connections by <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>
superusers. At most <A
HREF="runtime-config.html#GUC-MAX-CONNECTIONS"
>max_connections</A
>
connections can ever be active simultaneously. Whenever the
number of active concurrent connections is at least
<TT
CLASS="VARNAME"
>max_connections</TT
> minus
<TT
CLASS="VARNAME"
>superuser_reserved_connections</TT
>, new
connections will be accepted only for superusers.
</P
><P
> The default value is 2. The value must be less than the value of
<TT
CLASS="VARNAME"
>max_connections</TT
>. This parameter can only be
set at server start.
</P
></DD
><DT
><A
NAME="GUC-UNIX-SOCKET-DIRECTORY"
></A
><TT
CLASS="VARNAME"
>unix_socket_directory</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Specifies the directory of the Unix-domain socket on which the
server is to listen for
connections from client applications. The default is normally
<TT
CLASS="FILENAME"
>/tmp</TT
>, but can be changed at build time.
This parameter can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-UNIX-SOCKET-GROUP"
></A
><TT
CLASS="VARNAME"
>unix_socket_group</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Sets the owning group of the Unix-domain socket. (The owning
user of the socket is always the user that starts the
server.) In combination with the option
<TT
CLASS="VARNAME"
>unix_socket_permissions</TT
> this can be used as
an additional access control mechanism for Unix-domain connections.
By default this is the empty string, which uses the default
group for the current user. This option can only be set at
server start.
</P
></DD
><DT
><A
NAME="GUC-UNIX-SOCKET-PERMISSIONS"
></A
><TT
CLASS="VARNAME"
>unix_socket_permissions</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Sets the access permissions of the Unix-domain socket. Unix-domain
sockets use the usual Unix file system permission set.
The option value is expected to be a numeric mode
specification in the form accepted by the
<CODE
CLASS="FUNCTION"
>chmod</CODE
> and <CODE
CLASS="FUNCTION"
>umask</CODE
>
system calls. (To use the customary octal format the number
must start with a <TT
CLASS="LITERAL"
>0</TT
> (zero).)
</P
><P
> The default permissions are <TT
CLASS="LITERAL"
>0777</TT
>, meaning
anyone can connect. Reasonable alternatives are
<TT
CLASS="LITERAL"
>0770</TT
> (only user and group, see also
<TT
CLASS="VARNAME"
>unix_socket_group</TT
>) and <TT
CLASS="LITERAL"
>0700</TT
>
(only user). (Note that for a Unix-domain socket, only write
permission matters and so there is no point in setting or revoking
read or execute permissions.)
</P
><P
> This access control mechanism is independent of the one
described in <A
HREF="client-authentication.html"
>Chapter 19</A
>.
</P
><P
> This option can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-RENDEZVOUS-NAME"
></A
><TT
CLASS="VARNAME"
>rendezvous_name</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Specifies the <SPAN
CLASS="PRODUCTNAME"
>Rendezvous</SPAN
> broadcast
name. By default, the computer name is used, specified as an
empty string ''. This option is ignored if the server was not
compiled with <SPAN
CLASS="PRODUCTNAME"
>Rendezvous</SPAN
> support. This
option can only be set at server start.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-CONNECTION-SECURITY"
>16.4.2.2. Security and Authentication</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-AUTHENTICATION-TIMEOUT"
></A
><TT
CLASS="VARNAME"
>authentication_timeout</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Maximum time to complete client authentication, in seconds. If a
would-be client has not completed the authentication protocol in
this much time, the server breaks the connection. This prevents
hung clients from occupying a connection indefinitely. This
option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> file. The default is 60.
</P
></DD
><DT
><A
NAME="GUC-SSL"
></A
><TT
CLASS="VARNAME"
>ssl</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables <ACRONYM
CLASS="ACRONYM"
>SSL</ACRONYM
> connections. Please read
<A
HREF="ssl-tcp.html"
>Section 16.8</A
> before using this. The default
is off. This parameter can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-PASSWORD-ENCRYPTION"
></A
><TT
CLASS="VARNAME"
>password_encryption</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> When a password is specified in <A
HREF="sql-createuser.html"
><I
>CREATE USER</I
></A
> or
<A
HREF="sql-alteruser.html"
><I
>ALTER USER</I
></A
>
without writing either <TT
CLASS="LITERAL"
>ENCRYPTED</TT
> or
<TT
CLASS="LITERAL"
>UNENCRYPTED</TT
>, this option determines whether the
password is to be encrypted. The default is on (encrypt the
password).
</P
></DD
><DT
><A
NAME="GUC-KRB-SERVER-KEYFILE"
></A
><TT
CLASS="VARNAME"
>krb_server_keyfile</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Sets the location of the Kerberos server key file. See
<A
HREF="auth-methods.html#KERBEROS-AUTH"
>Section 19.2.3</A
> for details.
</P
></DD
><DT
><A
NAME="GUC-DB-USER-NAMESPACE"
></A
><TT
CLASS="VARNAME"
>db_user_namespace</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> This allows per-database user names. It is off by default.
</P
><P
> If this is on, you should create users as <TT
CLASS="LITERAL"
>username@dbname</TT
>.
When <TT
CLASS="LITERAL"
>username</TT
> is passed by a connecting client,
<TT
CLASS="LITERAL"
>@</TT
> and the database name is appended to the user
name and that database-specific user name is looked up by the
server. Note that when you create users with names containing
<TT
CLASS="LITERAL"
>@</TT
> within the SQL environment, you will need to
quote the user name.
</P
><P
> With this option enabled, you can still create ordinary global
users. Simply append <TT
CLASS="LITERAL"
>@</TT
> when specifying the user
name in the client. The <TT
CLASS="LITERAL"
>@</TT
> will be stripped off
before the user name is looked up by the server.
</P
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> This feature is intended as a temporary measure until a
complete solution is found. At that time, this option will
be removed.
</P
></BLOCKQUOTE
></DIV
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-RESOURCE"
>16.4.3. Resource Consumption</A
></H2
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-RESOURCE-MEMORY"
>16.4.3.1. Memory</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-SHARED-BUFFERS"
></A
><TT
CLASS="VARNAME"
>shared_buffers</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Sets the number of shared memory buffers used by the database
server. The default is typically 1000, but may be less if your
kernel settings will not support it (as determined during
<SPAN
CLASS="APPLICATION"
>initdb</SPAN
>). Each buffer is 8192 bytes, unless a
different value of <TT
CLASS="SYMBOL"
>BLCKSZ</TT
> was chosen when building
the server. This setting must be at least 16, as well as at
least twice the value of <A
HREF="runtime-config.html#GUC-MAX-CONNECTIONS"
>max_connections</A
>;
however, settings significantly higher than the minimum are
usually needed for good performance. Values of a few thousand
are recommended for production installations. This option can
only be set at server start.
</P
><P
> Increasing this parameter may cause <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>
to request more <SPAN
CLASS="SYSTEMITEM"
>System V</SPAN
> shared
memory than your operating system's default configuration
allows. See <A
HREF="kernel-resources.html#SYSVIPC"
>Section 16.5.1</A
> for information on how to
adjust those parameters, if necessary.
</P
></DD
><DT
><A
NAME="GUC-WORK-MEM"
></A
><TT
CLASS="VARNAME"
>work_mem</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Specifies the amount of memory to be used by internal sort operations
and hash tables before switching to temporary disk files. The value is
specified in kilobytes, and defaults to 1024 kilobytes (1 MB).
Note that for a complex query, several sort or hash operations might be
running in parallel; each one will be allowed to use as much memory
as this value specifies before it starts to put data into temporary
files. Also, several running sessions could be doing such operations
concurrently. So the total memory used could be many
times the value of <TT
CLASS="VARNAME"
>work_mem</TT
>; it is necessary to
keep this fact in mind when choosing the value. Sort operations are
used for <TT
CLASS="LITERAL"
>ORDER BY</TT
>, <TT
CLASS="LITERAL"
>DISTINCT</TT
>, and
merge joins.
Hash tables are used in hash joins, hash-based aggregation, and
hash-based processing of <TT
CLASS="LITERAL"
>IN</TT
> subqueries.
</P
></DD
><DT
><A
NAME="GUC-MAINTENANCE-WORK-MEM"
></A
><TT
CLASS="VARNAME"
>maintenance_work_mem</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Specifies the maximum amount of memory to be used in maintenance
operations, such as <TT
CLASS="COMMAND"
>VACUUM</TT
>, <TT
CLASS="COMMAND"
>CREATE
INDEX</TT
>, and <TT
CLASS="COMMAND"
>ALTER TABLE ADD FOREIGN KEY</TT
>.
The value is specified in kilobytes, and defaults to 16384 kilobytes
(16 MB). Since only one of these operations can be executed at
a time by a database session, and an installation normally doesn't
have very many of them happening concurrently, it's safe to set this
value significantly larger than <TT
CLASS="VARNAME"
>work_mem</TT
>. Larger
settings may improve performance for vacuuming and for restoring
database dumps.
</P
></DD
><DT
><A
NAME="GUC-MAX-STACK-DEPTH"
></A
><TT
CLASS="VARNAME"
>max_stack_depth</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Specifies the maximum safe depth of the server's execution stack.
The ideal setting for this parameter is the actual stack size limit
enforced by the kernel (as set by <TT
CLASS="LITERAL"
>ulimit -s</TT
> or local
equivalent), less a safety margin of a megabyte or so. The safety
margin is needed because the stack depth is not checked in every
routine in the server, but only in key potentially-recursive routines
such as expression evaluation. Setting the parameter higher than
the actual kernel limit will mean that a runaway recursive function
can crash an individual backend process. The default setting is
2048 KB (two megabytes), which is conservatively small and unlikely
to risk crashes. However, it may be too small to allow execution
of complex functions.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-RESOURCE-FSM"
>16.4.3.2. Free Space Map</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-MAX-FSM-PAGES"
></A
><TT
CLASS="VARNAME"
>max_fsm_pages</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Sets the maximum number of disk pages for which free space will
be tracked in the shared free-space map. Six bytes of shared memory
are consumed for each page slot. This setting must be more than
16 * <TT
CLASS="VARNAME"
>max_fsm_relations</TT
>. The default is 20000.
This option can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-MAX-FSM-RELATIONS"
></A
><TT
CLASS="VARNAME"
>max_fsm_relations</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Sets the maximum number of relations (tables and indexes) for which
free space will be tracked in the shared free-space map. Roughly
fifty bytes of shared memory are consumed for each slot.
The default is 1000.
This option can only be set at server start.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-RESOURCE-KERNEL"
>16.4.3.3. Kernel Resource Usage</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-MAX-FILES-PER-PROCESS"
></A
><TT
CLASS="VARNAME"
>max_files_per_process</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Sets the maximum number of simultaneously open files allowed to each
server subprocess. The default is 1000. If the kernel is enforcing
a safe per-process limit, you don't need to worry about this setting.
But on some platforms (notably, most BSD systems), the kernel will
allow individual processes to open many more files than the system
can really support when a large number of processes all try to open
that many files. If you find yourself seeing <SPAN
CLASS="QUOTE"
>"Too many open
files"</SPAN
> failures, try reducing this setting.
This option can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-PRELOAD-LIBRARIES"
></A
><TT
CLASS="VARNAME"
>preload_libraries</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> This variable specifies one or more shared libraries that are
to be preloaded at server start. A parameterless
initialization function can optionally be called for each
library. To specify that, add a colon and the name of the
initialization function after the library name. For example
<TT
CLASS="LITERAL"
>'$libdir/mylib:mylib_init'</TT
> would cause
<TT
CLASS="LITERAL"
>mylib</TT
> to be preloaded and <TT
CLASS="LITERAL"
>mylib_init</TT
>
to be executed. If more than one library is to be loaded,
separate their names with commas.
</P
><P
> If a specified library or initialization function is not found,
the server will fail to start.
</P
><P
> <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> procedural language
libraries may be preloaded in this way, typically by using the
syntax <TT
CLASS="LITERAL"
>'$libdir/plXXX:plXXX_init'</TT
> where
<TT
CLASS="LITERAL"
>XXX</TT
> is <TT
CLASS="LITERAL"
>pgsql</TT
>, <TT
CLASS="LITERAL"
>perl</TT
>,
<TT
CLASS="LITERAL"
>tcl</TT
>, or <TT
CLASS="LITERAL"
>python</TT
>.
</P
><P
> By preloading a shared library (and initializing it if
applicable), the library startup time is avoided when the
library is first used. However, the time to start each new
server process may increase slightly, even if that process never
uses the library. So this option is recommended only for
libraries that will be used in most sessions.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-RESOURCE-VACUUM-COST"
>16.4.3.4. Cost-Based Vacuum Delay</A
></H3
><P
> During the execution of <A
HREF="sql-vacuum.html"
><I
>VACUUM</I
></A
> and <A
HREF="sql-analyze.html"
><I
>ANALYZE</I
></A
> commands, the system maintains an
internal counter that keeps track of the estimated cost of the
various I/O operations that are performed. When the accumulated
cost reaches a limit (specified by
<TT
CLASS="VARNAME"
>vacuum_cost_limit</TT
>), the process performing
the operation will sleep for a while (specified by
<TT
CLASS="VARNAME"
>vacuum_cost_delay</TT
>). Then it will reset the
counter and continue execution.
</P
><P
> The intent of this feature is to allow administrators to reduce
the I/O impact of these commands on concurrent database
activity. There are many situations in which it is not very
important that maintenance commands like
<TT
CLASS="COMMAND"
>VACUUM</TT
> and <TT
CLASS="COMMAND"
>ANALYZE</TT
> finish
quickly; however, it is usually very important that these
commands do not significantly interfere with the ability of the
system to perform other database operations. Cost-based vacuum
delay provides a way for administrators to achieve this.
</P
><P
> This feature is disabled by default. To enable it, set the
<TT
CLASS="VARNAME"
>vacuum_cost_delay</TT
> variable to a nonzero
value.
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-VACUUM-COST-DELAY"
></A
><TT
CLASS="VARNAME"
>vacuum_cost_delay</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> The length of time, in milliseconds, that the process will sleep
when the cost limit has been exceeded.
The default value is 0, which disables the cost-based vacuum
delay feature. Positive values enable cost-based vacuuming.
Note that on many systems, the effective resolution
of sleep delays is 10 milliseconds; setting
<TT
CLASS="VARNAME"
>vacuum_cost_delay</TT
> to a value that is
not a multiple of 10 may have the same results as setting it
to the next higher multiple of 10.
</P
></DD
><DT
><A
NAME="GUC-VACUUM-COST-PAGE-HIT"
></A
><TT
CLASS="VARNAME"
>vacuum_cost_page_hit</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> The estimated cost for vacuuming a buffer found in the shared buffer
cache. It represents the cost to lock the buffer pool, lookup
the shared hash table and scan the content of the page. The
default value is 1.
</P
></DD
><DT
><A
NAME="GUC-VACUUM-COST-PAGE-MISS"
></A
><TT
CLASS="VARNAME"
>vacuum_cost_page_miss</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> The estimated cost for vacuuming a buffer that has to be read from
disk. This represents the effort to lock the buffer pool,
lookup the shared hash table, read the desired block in from
the disk and scan its content. The default value is 10.
</P
></DD
><DT
><A
NAME="GUC-VACUUM-COST-PAGE-DIRTY"
></A
><TT
CLASS="VARNAME"
>vacuum_cost_page_dirty</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> The estimated cost charged when vacuum modifies a block that was
previously clean. It represents the extra I/O required to
flush the dirty block out to disk again. The default value is
20.
</P
></DD
><DT
><A
NAME="GUC-VACUUM-COST-LIMIT"
></A
><TT
CLASS="VARNAME"
>vacuum_cost_limit</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> The accumulated cost that will cause the vacuuming process to sleep.
The default value is 200.
</P
></DD
></DL
></DIV
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> There are certain operations that hold critical locks and should
therefore complete as quickly as possible. Cost-based vacuum
delays do not occur during such operations. Therefore it is
possible that the cost accumulates far higher than the specified
limit. To avoid uselessly long delays in such cases, the actual
delay is calculated as <TT
CLASS="VARNAME"
>vacuum_cost_delay</TT
> *
<TT
CLASS="VARNAME"
>accumulated_balance</TT
> /
<TT
CLASS="VARNAME"
>vacuum_cost_limit</TT
> with a maximum of
<TT
CLASS="VARNAME"
>vacuum_cost_delay</TT
> * 4.
</P
></BLOCKQUOTE
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-RESOURCE-BACKGROUND-WRITER"
>16.4.3.5. Background Writer</A
></H3
><P
> Beginning in <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> 8.0, there is a separate server
process called the <I
CLASS="FIRSTTERM"
>background writer</I
>, whose sole function
is to issue writes of <SPAN
CLASS="QUOTE"
>"dirty"</SPAN
> shared buffers. The intent is
that server processes handling user queries should seldom or never have
to wait for a write to occur, because the background writer will do it.
This arrangement also reduces the performance penalty associated with
checkpoints. The background writer will continuously trickle out dirty
pages to disk, so that only a few pages will need to be forced out when
checkpoint time arrives, instead of the storm of dirty-buffer writes that
formerly occurred at each checkpoint. However there is a net overall
increase in I/O load, because where a repeatedly-dirtied page might
before have been written only once per checkpoint interval, the
background writer might write it several times in the same interval.
In most situations a continuous low load is preferable to periodic
spikes, but the parameters discussed in this section can be used to tune
the behavior for local needs.
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-BGWRITER-DELAY"
></A
><TT
CLASS="VARNAME"
>bgwriter_delay</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Specifies the delay between activity rounds for the
background writer. In each round the writer issues writes
for some number of dirty buffers (controllable by the
following parameters). The selected buffers will always be
the least recently used ones among the currently dirty
buffers. It then sleeps for <TT
CLASS="VARNAME"
>bgwriter_delay</TT
>
milliseconds, and repeats. The default value is 200. Note
that on many systems, the effective resolution of sleep
delays is 10 milliseconds; setting <TT
CLASS="VARNAME"
>bgwriter_delay</TT
>
to a value that is not a multiple of 10 may have the same
results as setting it to the next higher multiple of 10.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> file.
</P
></DD
><DT
><A
NAME="GUC-BGWRITER-PERCENT"
></A
><TT
CLASS="VARNAME"
>bgwriter_percent</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> In each round, no more than this percentage of the currently
dirty buffers will be written (rounding up any fraction to
the next whole number of buffers). The default value is
1. This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> file.
</P
></DD
><DT
><A
NAME="GUC-BGWRITER-MAXPAGES"
></A
><TT
CLASS="VARNAME"
>bgwriter_maxpages</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> In each round, no more than this many dirty buffers will be
written. The default value is 100. This option can only be
set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> file.
</P
></DD
></DL
></DIV
><P
> Smaller values of <TT
CLASS="VARNAME"
>bgwriter_percent</TT
> and
<TT
CLASS="VARNAME"
>bgwriter_maxpages</TT
> reduce the extra I/O load
caused by the background writer, but leave more work to be done
at checkpoint time. To reduce load spikes at checkpoints,
increase the values. To disable background writing entirely,
set <TT
CLASS="VARNAME"
>bgwriter_percent</TT
> and/or
<TT
CLASS="VARNAME"
>bgwriter_maxpages</TT
> to zero.
</P
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-WAL"
>16.4.4. Write Ahead Log</A
></H2
><P
> See also <A
HREF="wal-configuration.html"
>Section 25.2</A
> for details on WAL
tuning.
</P
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-WAL-SETTINGS"
>16.4.4.1. Settings</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-FSYNC"
></A
><TT
CLASS="VARNAME"
>fsync</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> If this option is on, the <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> server
will use the <CODE
CLASS="FUNCTION"
>fsync()</CODE
> system call in several places
to make sure that updates are physically written to disk. This
insures that a database cluster will recover to a
consistent state after an operating system or hardware crash.
</P
><P
> However, using <CODE
CLASS="FUNCTION"
>fsync()</CODE
> results in a
performance penalty: when a transaction is committed,
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> must wait for the
operating system to flush the write-ahead log to disk. When
<TT
CLASS="VARNAME"
>fsync</TT
> is disabled, the operating system is
allowed to do its best in buffering, ordering, and delaying
writes. This can result in significantly improved performance.
However, if the system crashes, the results of the last few
committed transactions may be lost in part or whole. In the
worst case, unrecoverable data corruption may occur.
(Crashes of the database server itself are <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>not</I
></SPAN
>
a risk factor here. Only an operating-system-level crash
creates a risk of corruption.)
</P
><P
> Due to the risks involved, there is no universally correct
setting for <TT
CLASS="VARNAME"
>fsync</TT
>. Some administrators
always disable <TT
CLASS="VARNAME"
>fsync</TT
>, while others only
turn it off for bulk loads, where there is a clear restart
point if something goes wrong, whereas some administrators
always leave <TT
CLASS="VARNAME"
>fsync</TT
> enabled. The default is
to enable <TT
CLASS="VARNAME"
>fsync</TT
>, for maximum reliability.
If you trust your operating system, your hardware, and your
utility company (or your battery backup), you can consider
disabling <TT
CLASS="VARNAME"
>fsync</TT
>.
</P
><P
> This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> file.
</P
></DD
><DT
><A
NAME="GUC-WAL-SYNC-METHOD"
></A
><TT
CLASS="VARNAME"
>wal_sync_method</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Method used for forcing WAL updates out to disk. Possible
values are
<TT
CLASS="LITERAL"
>fsync</TT
> (call <CODE
CLASS="FUNCTION"
>fsync()</CODE
> at each commit),
<TT
CLASS="LITERAL"
>fdatasync</TT
> (call <CODE
CLASS="FUNCTION"
>fdatasync()</CODE
> at each commit),
<TT
CLASS="LITERAL"
>fsync_writethrough</TT
> (call <CODE
CLASS="FUNCTION"
>_commit()</CODE
> at each commit on Windows),
<TT
CLASS="LITERAL"
>open_sync</TT
> (write WAL files with <CODE
CLASS="FUNCTION"
>open()</CODE
> option <TT
CLASS="SYMBOL"
>O_SYNC</TT
>), and
<TT
CLASS="LITERAL"
>open_datasync</TT
> (write WAL files with <CODE
CLASS="FUNCTION"
>open()</CODE
> option <TT
CLASS="SYMBOL"
>O_DSYNC</TT
>).
Not all of these choices are available on all platforms.
If <TT
CLASS="VARNAME"
>fsync</TT
> is off then this setting is irrelevant.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> file.
</P
></DD
><DT
><A
NAME="GUC-WAL-BUFFERS"
></A
><TT
CLASS="VARNAME"
>wal_buffers</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Number of disk-page buffers allocated in shared memory for WAL data.
The default is 8. The setting need only be large enough to hold
the amount of WAL data generated by one typical transaction.
This option can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-COMMIT-DELAY"
></A
><TT
CLASS="VARNAME"
>commit_delay</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Time delay between writing a commit record to the WAL buffer
and flushing the buffer out to disk, in microseconds. A
nonzero delay can allow multiple transactions to be committed
with only one <CODE
CLASS="FUNCTION"
>fsync()</CODE
> system call, if
system load is high enough that additional transactions become
ready to commit within the given interval. But the delay is
just wasted if no other transactions become ready to
commit. Therefore, the delay is only performed if at least
<TT
CLASS="VARNAME"
>commit_siblings</TT
> other transactions are
active at the instant that a server process has written its
commit record. The default is zero (no delay).
</P
></DD
><DT
><A
NAME="GUC-COMMIT-SIBLINGS"
></A
><TT
CLASS="VARNAME"
>commit_siblings</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Minimum number of concurrent open transactions to require
before performing the <TT
CLASS="VARNAME"
>commit_delay</TT
> delay. A larger
value makes it more probable that at least one other
transaction will become ready to commit during the delay
interval. The default is five.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-WAL-CHECKPOINTS"
>16.4.4.2. Checkpoints</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-CHECKPOINT-SEGMENTS"
></A
><TT
CLASS="VARNAME"
>checkpoint_segments</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Maximum distance between automatic WAL checkpoints, in log
file segments (each segment is normally 16 megabytes). The
default is three. This option can only be set at server start
or in the <TT
CLASS="FILENAME"
>postgresql.conf</TT
> file.
</P
></DD
><DT
><A
NAME="GUC-CHECKPOINT-TIMEOUT"
></A
><TT
CLASS="VARNAME"
>checkpoint_timeout</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Maximum time between automatic WAL checkpoints, in
seconds. The default is 300 seconds. This option can only be
set at server start or in the <TT
CLASS="FILENAME"
>postgresql.conf</TT
>
file.
</P
></DD
><DT
><A
NAME="GUC-CHECKPOINT-WARNING"
></A
><TT
CLASS="VARNAME"
>checkpoint_warning</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Write a message to the server log if checkpoints caused by
the filling of checkpoint segment files happen closer together
than this many seconds. The default is 30 seconds.
Zero turns off the warning.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-WAL-ARCHIVING"
>16.4.4.3. Archiving</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-ARCHIVE-COMMAND"
></A
><TT
CLASS="VARNAME"
>archive_command</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> The shell command to execute to archive a completed segment of
the WAL file series. If this is an empty string (the default),
WAL archiving is disabled. Any <TT
CLASS="LITERAL"
>%p</TT
> in the string is
replaced by the absolute path of the file to archive, and any
<TT
CLASS="LITERAL"
>%f</TT
> is replaced by the file name only. Use
<TT
CLASS="LITERAL"
>%%</TT
> to embed an actual <TT
CLASS="LITERAL"
>%</TT
> character in the
command. For more information see <A
HREF="backup-online.html#BACKUP-ARCHIVING-WAL"
>Section 22.3.1</A
>. This option can only be set at
server start or in the <TT
CLASS="FILENAME"
>postgresql.conf</TT
>
file.
</P
><P
> It is important for the command to return a zero exit status if
and only if it succeeds. Examples:
</P><PRE
CLASS="PROGRAMLISTING"
>archive_command = 'cp "%p" /mnt/server/archivedir/"%f"'
archive_command = 'copy "%p" /mnt/server/archivedir/"%f"' # Windows</PRE
><P>
</P
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-QUERY"
>16.4.5. Query Planning</A
></H2
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-QUERY-ENABLE"
>16.4.5.1. Planner Method Configuration</A
></H3
><P
> These configuration parameters provide a crude method of
influencing the query plans chosen by the query optimizer. If
the default plan chosen by the optimizer for a particular query
is not optimal, a temporary solution may be found by using one
of these configuration parameters to force the optimizer to
choose a different plan. Turning one of these settings off
permanently is seldom a good idea, however.
Better ways to improve the quality of the
plans chosen by the optimizer include adjusting the <A
HREF="runtime-config.html#RUNTIME-CONFIG-QUERY-CONSTANTS"
><I
>Planner Cost Constants</I
></A
>, running <A
HREF="sql-analyze.html"
><I
>ANALYZE</I
></A
> more
frequently, increasing the value of the <A
HREF="runtime-config.html#GUC-DEFAULT-STATISTICS-TARGET"
>default_statistics_target</A
> configuration parameter,
and increasing the amount of statistics collected for
specific columns using <TT
CLASS="COMMAND"
>ALTER TABLE SET
STATISTICS</TT
>.
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-ENABLE-HASHAGG"
></A
><TT
CLASS="VARNAME"
>enable_hashagg</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables or disables the query planner's use of hashed
aggregation plan types. The default is on.
</P
></DD
><DT
><A
NAME="GUC-ENABLE-HASHJOIN"
></A
><TT
CLASS="VARNAME"
>enable_hashjoin</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables or disables the query planner's use of hash-join plan
types. The default is on.
</P
></DD
><DT
><A
NAME="GUC-ENABLE-INDEXSCAN"
></A
><TT
CLASS="VARNAME"
>enable_indexscan</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables or disables the query planner's use of index-scan plan
types. The default is on.
</P
></DD
><DT
><A
NAME="GUC-ENABLE-MERGEJOIN"
></A
><TT
CLASS="VARNAME"
>enable_mergejoin</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables or disables the query planner's use of merge-join plan
types. The default is on.
</P
></DD
><DT
><A
NAME="GUC-ENABLE-NESTLOOP"
></A
><TT
CLASS="VARNAME"
>enable_nestloop</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables or disables the query planner's use of nested-loop join
plans. It's not possible to suppress nested-loop joins entirely,
but turning this variable off discourages the planner from using
one if there are other methods available. The default is
on.
</P
></DD
><DT
><A
NAME="GUC-ENABLE-SEQSCAN"
></A
><TT
CLASS="VARNAME"
>enable_seqscan</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables or disables the query planner's use of sequential scan
plan types. It's not possible to suppress sequential scans
entirely, but turning this variable off discourages the planner
from using one if there are other methods available. The
default is on.
</P
></DD
><DT
><A
NAME="GUC-ENABLE-SORT"
></A
><TT
CLASS="VARNAME"
>enable_sort</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables or disables the query planner's use of explicit sort
steps. It's not possible to suppress explicit sorts entirely,
but turning this variable off discourages the planner from
using one if there are other methods available. The default
is on.
</P
></DD
><DT
><A
NAME="GUC-ENABLE-TIDSCAN"
></A
><TT
CLASS="VARNAME"
>enable_tidscan</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables or disables the query planner's use of <ACRONYM
CLASS="ACRONYM"
>TID</ACRONYM
>
scan plan types. The default is on.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-QUERY-CONSTANTS"
>16.4.5.2. Planner Cost Constants</A
></H3
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> Unfortunately, there is no well-defined method for determining
ideal values for the family of <SPAN
CLASS="QUOTE"
>"cost"</SPAN
> variables that
appear below. You are encouraged to experiment and share
your findings.
</P
></BLOCKQUOTE
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-EFFECTIVE-CACHE-SIZE"
></A
><TT
CLASS="VARNAME"
>effective_cache_size</TT
> (<TT
CLASS="TYPE"
>floating point</TT
>)</DT
><DD
><P
> Sets the planner's assumption about the effective size of the
disk cache that is available to a single index scan. This is
factored into estimates of the cost of using an index; a higher
value makes it more likely index scans will be used, a lower
value makes it more likely sequential scans will be used. When
setting this parameter you should consider both
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>'s shared buffers and the
portion of the kernel's disk cache that will be used for
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> data files. Also, take into
account the expected number of concurrent queries using different
indexes, since they will have to share the available space.
This parameter has no effect on the size of shared memory
allocated by PostgreSQL, nor does it reserve kernel disk cache;
it is used only for estimation purposes.
The value is measured in disk pages, which are
normally 8192 bytes each. The default is 1000.
</P
></DD
><DT
><A
NAME="GUC-RANDOM-PAGE-COST"
></A
><TT
CLASS="VARNAME"
>random_page_cost</TT
> (<TT
CLASS="TYPE"
>floating point</TT
>)</DT
><DD
><P
> Sets the planner's estimate of the cost of a
nonsequentially fetched disk page. This is measured as a
multiple of the cost of a sequential page fetch. A higher
value makes it more likely a sequential scan will be used, a
lower value makes it more likely an index scan will be
used. The default is four.
</P
></DD
><DT
><A
NAME="GUC-CPU-TUPLE-COST"
></A
><TT
CLASS="VARNAME"
>cpu_tuple_cost</TT
> (<TT
CLASS="TYPE"
>floating point</TT
>)</DT
><DD
><P
> Sets the planner's estimate of the cost of processing
each row during a query. This is measured as a fraction of
the cost of a sequential page fetch. The default is 0.01.
</P
></DD
><DT
><A
NAME="GUC-CPU-INDEX-TUPLE-COST"
></A
><TT
CLASS="VARNAME"
>cpu_index_tuple_cost</TT
> (<TT
CLASS="TYPE"
>floating point</TT
>)</DT
><DD
><P
> Sets the planner's estimate of the cost of processing
each index row during an index scan. This is measured as a
fraction of the cost of a sequential page fetch. The default
is 0.001.
</P
></DD
><DT
><A
NAME="GUC-CPU-OPERATOR-COST"
></A
><TT
CLASS="VARNAME"
>cpu_operator_cost</TT
> (<TT
CLASS="TYPE"
>floating point</TT
>)</DT
><DD
><P
> Sets the planner's estimate of the cost of processing each
operator in a <TT
CLASS="LITERAL"
>WHERE</TT
> clause. This is measured as a fraction of
the cost of a sequential page fetch. The default is 0.0025.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-QUERY-GEQO"
>16.4.5.3. Genetic Query Optimizer</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-GEQO"
></A
><TT
CLASS="VARNAME"
>geqo</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables or disables genetic query optimization, which is an
algorithm that attempts to do query planning without
exhaustive searching. This is on by default. The
<TT
CLASS="VARNAME"
>geqo_threshold</TT
> variable provides a more
granular way to disable GEQO for certain classes of queries.
</P
></DD
><DT
><A
NAME="GUC-GEQO-THRESHOLD"
></A
><TT
CLASS="VARNAME"
>geqo_threshold</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Use genetic query optimization to plan queries with at least
this many <TT
CLASS="LITERAL"
>FROM</TT
> items involved. (Note that an outer
<TT
CLASS="LITERAL"
>JOIN</TT
> construct counts as only one <TT
CLASS="LITERAL"
>FROM</TT
>
item.) The default is 12. For simpler queries it is usually best
to use the deterministic, exhaustive planner, but for queries with
many tables the deterministic planner takes too long.
</P
></DD
><DT
><A
NAME="GUC-GEQO-EFFORT"
></A
><TT
CLASS="VARNAME"
>geqo_effort</TT
>
(<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Controls the trade off between planning time and query plan
efficiency in GEQO. This variable must be an integer in the
range from 1 to 10. The default value is 5. Larger values
increase the time spent doing query planning, but also
increase the likelihood that an efficient query plan will be
chosen.
</P
><P
> <TT
CLASS="VARNAME"
>geqo_effort</TT
> doesn't actually do anything
directly; it is only used to compute the default values for
the other variables that influence GEQO behavior (described
below). If you prefer, you can set the other parameters by
hand instead.
</P
></DD
><DT
><A
NAME="GUC-GEQO-POOL-SIZE"
></A
><TT
CLASS="VARNAME"
>geqo_pool_size</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Controls the pool size used by GEQO. The pool size is the
number of individuals in the genetic population. It must be
at least two, and useful values are typically 100 to 1000. If
it is set to zero (the default setting) then a suitable
default is chosen based on <TT
CLASS="VARNAME"
>geqo_effort</TT
> and
the number of tables in the query.
</P
></DD
><DT
><A
NAME="GUC-GEQO-GENERATIONS"
></A
><TT
CLASS="VARNAME"
>geqo_generations</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Controls the number of generations used by GEQO. Generations
specifies the number of iterations of the algorithm. It must
be at least one, and useful values are in the same range as
the pool size. If it is set to zero (the default setting)
then a suitable default is chosen based on
<TT
CLASS="VARNAME"
>geqo_pool_size</TT
>.
</P
></DD
><DT
><A
NAME="GUC-GEQO-SELECTION-BIAS"
></A
><TT
CLASS="VARNAME"
>geqo_selection_bias</TT
> (<TT
CLASS="TYPE"
>floating point</TT
>)</DT
><DD
><P
> Controls the selection bias used by GEQO. The selection bias
is the selective pressure within the population. Values can be
from 1.50 to 2.00; the latter is the default.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-QUERY-OTHER"
>16.4.5.4. Other Planner Options</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-DEFAULT-STATISTICS-TARGET"
></A
><TT
CLASS="VARNAME"
>default_statistics_target</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Sets the default statistics target for table columns that have
not had a column-specific target set via <TT
CLASS="COMMAND"
>ALTER TABLE
SET STATISTICS</TT
>. Larger values increase the time needed to
do <TT
CLASS="COMMAND"
>ANALYZE</TT
>, but may improve the quality of the
planner's estimates. The default is 10. For more information
on the use of statistics by the <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>
query planner, refer to <A
HREF="planner-stats.html"
>Section 13.2</A
>.
</P
></DD
><DT
><A
NAME="GUC-FROM-COLLAPSE-LIMIT"
></A
><TT
CLASS="VARNAME"
>from_collapse_limit</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> The planner will merge sub-queries into upper queries if the
resulting <TT
CLASS="LITERAL"
>FROM</TT
> list would have no more than
this many items. Smaller values reduce planning time but may
yield inferior query plans. The default is 8. It is usually
wise to keep this less than <A
HREF="runtime-config.html#GUC-GEQO-THRESHOLD"
>geqo_threshold</A
>.
</P
></DD
><DT
><A
NAME="GUC-JOIN-COLLAPSE-LIMIT"
></A
><TT
CLASS="VARNAME"
>join_collapse_limit</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> The planner will rewrite explicit inner <TT
CLASS="LITERAL"
>JOIN</TT
>
constructs into lists of <TT
CLASS="LITERAL"
>FROM</TT
> items whenever a
list of no more than this many items in total would
result. Prior to <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> 7.4, joins
specified via the <TT
CLASS="LITERAL"
>JOIN</TT
> construct would
never be reordered by the query planner. The query planner has
subsequently been improved so that inner joins written in this
form can be reordered; this configuration parameter controls
the extent to which this reordering is performed.
</P><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> At present, the order of outer joins specified via the
<TT
CLASS="LITERAL"
>JOIN</TT
> construct is never adjusted by the query
planner; therefore, <TT
CLASS="VARNAME"
>join_collapse_limit</TT
> has no
effect on this behavior. The planner may be improved to
reorder some classes of outer joins in a future release of
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>.
</P
></BLOCKQUOTE
></DIV
><P>
</P
><P
> By default, this variable is set the same as
<TT
CLASS="VARNAME"
>from_collapse_limit</TT
>, which is appropriate
for most uses. Setting it to 1 prevents any reordering of
inner <TT
CLASS="LITERAL"
>JOIN</TT
>s. Thus, the explicit join order
specified in the query will be the actual order in which the
relations are joined. The query planner does not always choose
the optimal join order; advanced users may elect to
temporarily set this variable to 1, and then specify the join
order they desire explicitly. Another consequence of setting
this variable to 1 is that the query planner will behave more
like the <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> 7.3 query
planner, which some users might find useful for backward
compatibility reasons.
</P
><P
> Setting this variable to a value between 1 and
<TT
CLASS="VARNAME"
>from_collapse_limit</TT
> might be useful to
trade off planning time against the quality of the chosen plan
(higher values produce better plans).
</P
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-LOGGING"
>16.4.6. Error Reporting and Logging</A
></H2
><A
NAME="AEN17556"
></A
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-LOGGING-WHERE"
>16.4.6.1. Where to log</A
></H3
><A
NAME="AEN17560"
></A
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-LOG-DESTINATION"
></A
><TT
CLASS="VARNAME"
>log_destination</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> supports several methods
for logging server messages, including
<SPAN
CLASS="SYSTEMITEM"
>stderr</SPAN
> and
<SPAN
CLASS="SYSTEMITEM"
>syslog</SPAN
>. On Windows,
<SPAN
CLASS="SYSTEMITEM"
>eventlog</SPAN
> is also supported. Set this
option to a list of desired log destinations separated by
commas. The default is to log to <SPAN
CLASS="SYSTEMITEM"
>stderr</SPAN
>
only.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
</P
></DD
><DT
><A
NAME="GUC-REDIRECT-STDERR"
></A
><TT
CLASS="VARNAME"
>redirect_stderr</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> This option allows messages sent to <SPAN
CLASS="APPLICATION"
>stderr</SPAN
> to be
captured and redirected into log files.
This option, in combination with logging to <SPAN
CLASS="APPLICATION"
>stderr</SPAN
>,
is often more useful than
logging to <SPAN
CLASS="APPLICATION"
>syslog</SPAN
>, since some types of messages
may not appear in <SPAN
CLASS="APPLICATION"
>syslog</SPAN
> output (a common example
is dynamic-linker failure messages).
This option can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-LOG-DIRECTORY"
></A
><TT
CLASS="VARNAME"
>log_directory</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> When <TT
CLASS="VARNAME"
>redirect_stderr</TT
> is enabled, this option
determines the directory in which log files will be created.
It may be specified as an absolute path, or relative to the
cluster data directory.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
</P
></DD
><DT
><A
NAME="GUC-LOG-FILENAME"
></A
><TT
CLASS="VARNAME"
>log_filename</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> When <TT
CLASS="VARNAME"
>redirect_stderr</TT
> is enabled, this option
sets the file names of the created log files. The value
is treated as a <SPAN
CLASS="SYSTEMITEM"
>strftime</SPAN
> pattern,
so <TT
CLASS="LITERAL"
>%</TT
>-escapes
can be used to specify time-varying file names.
If no <TT
CLASS="LITERAL"
>%</TT
>-escapes are present,
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> will
append the epoch of the new log file's open time. For example,
if <TT
CLASS="VARNAME"
>log_filename</TT
> were <TT
CLASS="LITERAL"
>server_log</TT
>, then the
chosen file name would be <TT
CLASS="LITERAL"
>server_log.1093827753</TT
>
for a log starting at Sun Aug 29 19:02:33 2004 MST.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
</P
></DD
><DT
><A
NAME="GUC-LOG-ROTATION-AGE"
></A
><TT
CLASS="VARNAME"
>log_rotation_age</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> When <TT
CLASS="VARNAME"
>redirect_stderr</TT
> is enabled, this option
determines the maximum lifetime of an individual log file.
After this many minutes have elapsed, a new log file will
be created. Set to zero to disable time-based creation of
new log files.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
</P
></DD
><DT
><A
NAME="GUC-LOG-ROTATION-SIZE"
></A
><TT
CLASS="VARNAME"
>log_rotation_size</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> When <TT
CLASS="VARNAME"
>redirect_stderr</TT
> is enabled, this option
determines the maximum size of an individual log file.
After this many kilobytes have been emitted into a log file,
a new log file will be created. Set to zero to disable size-based
creation of new log files.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
</P
></DD
><DT
><A
NAME="GUC-LOG-TRUNCATE-ON-ROTATION"
></A
><TT
CLASS="VARNAME"
>log_truncate_on_rotation</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> When <TT
CLASS="VARNAME"
>redirect_stderr</TT
> is enabled, this option will cause
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> to truncate (overwrite),
rather than append to, any existing log file of the same name.
However, truncation will occur only when a new file is being opened
due to time-based rotation, not during server startup or size-based
rotation. When false, pre-existing files will be appended to in
all cases. For example, using this option in combination with
a <TT
CLASS="VARNAME"
>log_filename</TT
> like <TT
CLASS="LITERAL"
>postgresql-%H.log</TT
>
would result in generating twenty-four hourly log files and then
cyclically overwriting them.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
</P
><P
> Example: To keep 7 days of logs, one log file per day named
<TT
CLASS="LITERAL"
>server_log.Mon</TT
>, <TT
CLASS="LITERAL"
>server_log.Tue</TT
>,
etc, and automatically overwrite last week's log with this week's log,
set <TT
CLASS="VARNAME"
>log_filename</TT
> to <TT
CLASS="LITERAL"
>server_log.%a</TT
>,
<TT
CLASS="VARNAME"
>log_truncate_on_rotation</TT
> to <TT
CLASS="LITERAL"
>true</TT
>, and
<TT
CLASS="VARNAME"
>log_rotation_age</TT
> to <TT
CLASS="LITERAL"
>1440</TT
>.
</P
><P
> Example: To keep 24 hours of logs, one log file per hour, but
also rotate sooner if the log file size exceeds 1GB, set
<TT
CLASS="VARNAME"
>log_filename</TT
> to <TT
CLASS="LITERAL"
>server_log.%H%M</TT
>,
<TT
CLASS="VARNAME"
>log_truncate_on_rotation</TT
> to <TT
CLASS="LITERAL"
>true</TT
>,
<TT
CLASS="VARNAME"
>log_rotation_age</TT
> to <TT
CLASS="LITERAL"
>60</TT
>, and
<TT
CLASS="VARNAME"
>log_rotation_size</TT
> to <TT
CLASS="LITERAL"
>1000000</TT
>.
Including <TT
CLASS="LITERAL"
>%M</TT
> in <TT
CLASS="VARNAME"
>log_filename</TT
> allows
any size-driven rotations that may occur to select a filename
different from the hour's initial filename.
</P
></DD
><DT
><A
NAME="GUC-SYSLOG-FACILITY"
></A
><TT
CLASS="VARNAME"
>syslog_facility</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> When logging to <SPAN
CLASS="APPLICATION"
>syslog</SPAN
> is enabled, this option
determines the <SPAN
CLASS="APPLICATION"
>syslog</SPAN
>
<SPAN
CLASS="QUOTE"
>"facility"</SPAN
> to be used. You may choose
from <TT
CLASS="LITERAL"
>LOCAL0</TT
>, <TT
CLASS="LITERAL"
>LOCAL1</TT
>,
<TT
CLASS="LITERAL"
>LOCAL2</TT
>, <TT
CLASS="LITERAL"
>LOCAL3</TT
>, <TT
CLASS="LITERAL"
>LOCAL4</TT
>,
<TT
CLASS="LITERAL"
>LOCAL5</TT
>, <TT
CLASS="LITERAL"
>LOCAL6</TT
>, <TT
CLASS="LITERAL"
>LOCAL7</TT
>;
the default is <TT
CLASS="LITERAL"
>LOCAL0</TT
>. See also the
documentation of your system's
<SPAN
CLASS="APPLICATION"
>syslog</SPAN
> daemon.
This option can only be set at server start.
</P
></DD
><DT
><A
NAME="GUC-SYSLOG-IDENT"
></A
><TT
CLASS="VARNAME"
>syslog_ident</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> When logging to <SPAN
CLASS="APPLICATION"
>syslog</SPAN
> is enabled, this option
determines the program name used to identify
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> messages in
<SPAN
CLASS="APPLICATION"
>syslog</SPAN
> logs. The default is
<TT
CLASS="LITERAL"
>postgres</TT
>.
This option can only be set at server start.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-LOGGING-WHEN"
>16.4.6.2. When To Log</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-CLIENT-MIN-MESSAGES"
></A
><TT
CLASS="VARNAME"
>client_min_messages</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Controls which message levels are sent to the client.
Valid values are <TT
CLASS="LITERAL"
>DEBUG5</TT
>,
<TT
CLASS="LITERAL"
>DEBUG4</TT
>, <TT
CLASS="LITERAL"
>DEBUG3</TT
>, <TT
CLASS="LITERAL"
>DEBUG2</TT
>,
<TT
CLASS="LITERAL"
>DEBUG1</TT
>, <TT
CLASS="LITERAL"
>LOG</TT
>, <TT
CLASS="LITERAL"
>NOTICE</TT
>,
<TT
CLASS="LITERAL"
>WARNING</TT
>, and <TT
CLASS="LITERAL"
>ERROR</TT
>. Each level
includes all the levels that follow it. The later the level,
the fewer messages are sent. The default is
<TT
CLASS="LITERAL"
>NOTICE</TT
>. Note that <TT
CLASS="LITERAL"
>LOG</TT
> has a different
rank here than in <TT
CLASS="VARNAME"
>log_min_messages</TT
>.
</P
></DD
><DT
><A
NAME="GUC-LOG-MIN-MESSAGES"
></A
><TT
CLASS="VARNAME"
>log_min_messages</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Controls which message levels are written to the server log.
Valid values are <TT
CLASS="LITERAL"
>DEBUG5</TT
>, <TT
CLASS="LITERAL"
>DEBUG4</TT
>,
<TT
CLASS="LITERAL"
>DEBUG3</TT
>, <TT
CLASS="LITERAL"
>DEBUG2</TT
>, <TT
CLASS="LITERAL"
>DEBUG1</TT
>,
<TT
CLASS="LITERAL"
>INFO</TT
>, <TT
CLASS="LITERAL"
>NOTICE</TT
>, <TT
CLASS="LITERAL"
>WARNING</TT
>,
<TT
CLASS="LITERAL"
>ERROR</TT
>, <TT
CLASS="LITERAL"
>LOG</TT
>, <TT
CLASS="LITERAL"
>FATAL</TT
>, and
<TT
CLASS="LITERAL"
>PANIC</TT
>. Each level includes all the levels that
follow it. The later the level, the fewer messages are sent
to the log. The default is <TT
CLASS="LITERAL"
>NOTICE</TT
>. Note that
<TT
CLASS="LITERAL"
>LOG</TT
> has a different rank here than in
<TT
CLASS="VARNAME"
>client_min_messages</TT
>.
Only superusers can change this setting.
</P
></DD
><DT
><A
NAME="GUC-LOG-ERROR-VERBOSITY"
></A
><TT
CLASS="VARNAME"
>log_error_verbosity</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Controls the amount of detail written in the server log for each
message that is logged. Valid values are <TT
CLASS="LITERAL"
>TERSE</TT
>,
<TT
CLASS="LITERAL"
>DEFAULT</TT
>, and <TT
CLASS="LITERAL"
>VERBOSE</TT
>, each adding more
fields to displayed messages.
Only superusers can change this setting.
</P
></DD
><DT
><A
NAME="GUC-LOG-MIN-ERROR-STATEMENT"
></A
><TT
CLASS="VARNAME"
>log_min_error_statement</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Controls whether or not the SQL statement that causes an error
condition will also be recorded in the server log. All SQL
statements that cause an error of the specified level or
higher are logged. The default is
<TT
CLASS="LITERAL"
>PANIC</TT
> (effectively turning this feature
off for normal use). Valid values are <TT
CLASS="LITERAL"
>DEBUG5</TT
>,
<TT
CLASS="LITERAL"
>DEBUG4</TT
>, <TT
CLASS="LITERAL"
>DEBUG3</TT
>,
<TT
CLASS="LITERAL"
>DEBUG2</TT
>, <TT
CLASS="LITERAL"
>DEBUG1</TT
>,
<TT
CLASS="LITERAL"
>INFO</TT
>, <TT
CLASS="LITERAL"
>NOTICE</TT
>,
<TT
CLASS="LITERAL"
>WARNING</TT
>, <TT
CLASS="LITERAL"
>ERROR</TT
>,
<TT
CLASS="LITERAL"
>FATAL</TT
>, and <TT
CLASS="LITERAL"
>PANIC</TT
>. For
example, if you set this to <TT
CLASS="LITERAL"
>ERROR</TT
> then all
SQL statements causing errors, fatal errors, or panics will be
logged. Enabling this option can be helpful in tracking down
the source of any errors that appear in the server log.
Only superusers can change this setting.
</P
></DD
><DT
><A
NAME="GUC-LOG-MIN-DURATION-STATEMENT"
></A
><TT
CLASS="VARNAME"
>log_min_duration_statement</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Sets a minimum statement execution time (in milliseconds)
that causes a statement to be logged. All SQL statements
that run for the time specified or longer will be logged with
their duration. Setting this to zero will print
all queries and their durations. Minus-one (the default)
disables the feature. For example, if you set it to
<TT
CLASS="LITERAL"
>250</TT
> then all SQL statements that run 250ms
or longer will be logged. Enabling this option can be
useful in tracking down unoptimized queries in your applications.
Only superusers can change this setting.
</P
></DD
><DT
><A
NAME="GUC-SILENT-MODE"
></A
><TT
CLASS="VARNAME"
>silent_mode</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Runs the server silently. If this option is set, the server
will automatically run in background and any controlling
terminals are disassociated (same effect as
<TT
CLASS="COMMAND"
>postmaster</TT
>'s <TT
CLASS="OPTION"
>-S</TT
> option).
The server's standard output and standard error are redirected
to <TT
CLASS="LITERAL"
>/dev/null</TT
>, so any messages sent to them will be lost.
Unless <SPAN
CLASS="APPLICATION"
>syslog</SPAN
> logging is selected or
<TT
CLASS="VARNAME"
>redirect_stderr</TT
> is enabled, using this option
is discouraged because it makes it impossible to see error messages.
</P
></DD
></DL
></DIV
><P
> Here is a list of the various message severity levels used in
these settings:
<P
></P
></P><DIV
CLASS="VARIABLELIST"
><DL
><DT
><TT
CLASS="LITERAL"
>DEBUG[1-5]</TT
></DT
><DD
><P
> Provides information for use by developers.
</P
></DD
><DT
><TT
CLASS="LITERAL"
>INFO</TT
></DT
><DD
><P
> Provides information implicitly requested by the user,
e.g., during <TT
CLASS="COMMAND"
>VACUUM VERBOSE</TT
>.
</P
></DD
><DT
><TT
CLASS="LITERAL"
>NOTICE</TT
></DT
><DD
><P
> Provides information that may be helpful to users, e.g.,
truncation of long identifiers and the creation of indexes as part
of primary keys.
</P
></DD
><DT
><TT
CLASS="LITERAL"
>WARNING</TT
></DT
><DD
><P
> Provides warnings to the user, e.g., <TT
CLASS="COMMAND"
>COMMIT</TT
>
outside a transaction block.
</P
></DD
><DT
><TT
CLASS="LITERAL"
>ERROR</TT
></DT
><DD
><P
> Reports an error that caused the current command to abort.
</P
></DD
><DT
><TT
CLASS="LITERAL"
>LOG</TT
></DT
><DD
><P
> Reports information of interest to administrators, e.g.,
checkpoint activity.
</P
></DD
><DT
><TT
CLASS="LITERAL"
>FATAL</TT
></DT
><DD
><P
> Reports an error that caused the current session to abort.
</P
></DD
><DT
><TT
CLASS="LITERAL"
>PANIC</TT
></DT
><DD
><P
> Reports an error that caused all sessions to abort.
</P
></DD
></DL
></DIV
><P>
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-LOGGING-WHAT"
>16.4.6.3. What To Log</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><TT
CLASS="VARNAME"
>debug_print_parse</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>debug_print_rewritten</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>debug_print_plan</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>debug_pretty_print</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> These options enable various debugging output to be emitted.
For each executed query, they print
the resulting parse tree, the query rewriter output, or the
execution plan. <TT
CLASS="VARNAME"
>debug_pretty_print</TT
> indents
these displays to produce a more readable but much longer
output format. <TT
CLASS="VARNAME"
>client_min_messages</TT
> or
<TT
CLASS="VARNAME"
>log_min_messages</TT
> must be
<TT
CLASS="LITERAL"
>DEBUG1</TT
> or lower to actually send this output
to the client or the server log, respectively.
These options are off by default.
</P
></DD
><DT
><A
NAME="GUC-LOG-CONNECTIONS"
></A
><TT
CLASS="VARNAME"
>log_connections</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> This outputs a line to the server log detailing each successful
connection. This is off by default, although it is probably very
useful. This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
</P
></DD
><DT
><A
NAME="GUC-LOG-DISCONNECTIONS"
></A
><TT
CLASS="VARNAME"
>log_disconnections</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> This outputs a line in the server log similar to
<TT
CLASS="VARNAME"
>log_connections</TT
> but at session termination,
and includes the duration of the session. This is off by
default. This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
</P
></DD
><DT
><A
NAME="GUC-LOG-DURATION"
></A
><TT
CLASS="VARNAME"
>log_duration</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Causes the duration of every completed statement which satisfies
<TT
CLASS="VARNAME"
>log_statement</TT
> to be logged. When using this option,
if you are not using <SPAN
CLASS="APPLICATION"
>syslog</SPAN
>, it is recommended
that you log the PID or session ID using <TT
CLASS="VARNAME"
>log_line_prefix</TT
>
so that you can link the statement to the
duration using the process ID or session ID. The default is off.
Only superusers can change this setting.
</P
></DD
><DT
><A
NAME="GUC-LOG-LINE-PREFIX"
></A
><TT
CLASS="VARNAME"
>log_line_prefix</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> This is a <CODE
CLASS="FUNCTION"
>printf</CODE
>-style string that is output at the
beginning of each log line. The default is an empty string.
Each recognized escape is replaced as outlined
below - anything else that looks like an escape is ignored. Other
characters are copied straight to the log line. Some escapes are
only recognised by session processes, and do not apply to
background processes such as the postmaster. <SPAN
CLASS="APPLICATION"
>Syslog</SPAN
>
produces its own
time stamp and process ID information, so you probably do not want to
use those escapes if you are using <SPAN
CLASS="APPLICATION"
>syslog</SPAN
>.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
<DIV
CLASS="INFORMALTABLE"
><P
></P
><A
NAME="AEN17941"
></A
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><COL><THEAD
><TR
><TH
>Escape</TH
><TH
>Effect</TH
><TH
>Session only</TH
></TR
></THEAD
><TBODY
><TR
><TD
><TT
CLASS="LITERAL"
>%u</TT
></TD
><TD
>User name</TD
><TD
>yes</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%d</TT
></TD
><TD
>Database name</TD
><TD
>yes</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%r</TT
></TD
><TD
>Remote host name or IP address, and remote port</TD
><TD
>yes</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%p</TT
></TD
><TD
>Process ID</TD
><TD
>no</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%t</TT
></TD
><TD
>Time stamp</TD
><TD
>no</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%i</TT
></TD
><TD
>Command tag: This is the command that generated the log line.</TD
><TD
>yes</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%c</TT
></TD
><TD
>Session ID: A unique identifier for each session.
It is 2 4-byte hexadecimal numbers (without leading zeros)
separated by a dot. The numbers
are the session start time and the process ID, so this can also
be used as a space saving way of printing these items.</TD
><TD
>yes</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%l</TT
></TD
><TD
>Number of the log line for each process, starting at 1</TD
><TD
>no</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%s</TT
></TD
><TD
>Session start time stamp</TD
><TD
>yes</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%x</TT
></TD
><TD
>Transaction ID</TD
><TD
>yes</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%q</TT
></TD
><TD
>Does not produce any output, but tells non-session
processes to stop at this point in the string. Ignored by
session processes.</TD
><TD
>no</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>%%</TT
></TD
><TD
>Literal <TT
CLASS="LITERAL"
>%</TT
></TD
><TD
>no</TD
></TR
></TBODY
></TABLE
><P
></P
></DIV
>
</P
></DD
><DT
><A
NAME="GUC-LOG-STATEMENT"
></A
><TT
CLASS="VARNAME"
>log_statement</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Controls which SQL statements are logged. Valid values are
<TT
CLASS="LITERAL"
>none</TT
>, <TT
CLASS="LITERAL"
>ddl</TT
>, <TT
CLASS="LITERAL"
>mod</TT
>, and
<TT
CLASS="LITERAL"
>all</TT
>. <TT
CLASS="LITERAL"
>ddl</TT
> logs all data definition
commands like <TT
CLASS="LITERAL"
>CREATE</TT
>, <TT
CLASS="LITERAL"
>ALTER</TT
>, and
<TT
CLASS="LITERAL"
>DROP</TT
> commands. <TT
CLASS="LITERAL"
>mod</TT
> logs all
<TT
CLASS="LITERAL"
>ddl</TT
> statements, plus <TT
CLASS="LITERAL"
>INSERT</TT
>,
<TT
CLASS="LITERAL"
>UPDATE</TT
>, <TT
CLASS="LITERAL"
>DELETE</TT
>, <TT
CLASS="LITERAL"
>TRUNCATE</TT
>,
and <TT
CLASS="LITERAL"
>COPY FROM</TT
>. <TT
CLASS="LITERAL"
>PREPARE</TT
> and
<TT
CLASS="LITERAL"
>EXPLAIN ANALYZE</TT
> statements are also logged if their
contained command is of an appropriate type.
</P
><P
> The default is <TT
CLASS="LITERAL"
>none</TT
>. Only superusers can change this
setting.
</P
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> The <TT
CLASS="COMMAND"
>EXECUTE</TT
> statement is not considered a
<TT
CLASS="LITERAL"
>ddl</TT
> or <TT
CLASS="LITERAL"
>mod</TT
> statement. When it is logged,
only the name of the prepared statement is reported, not the
actual prepared statement.
</P
><P
> When a function is defined in the
<SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
>server-side language, any queries
executed by the function will only be logged the first time that the
function is invoked in a particular session. This is because
<SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> keeps a cache of the
query plans produced for the SQL statements in the function.
</P
></BLOCKQUOTE
></DIV
></DD
><DT
><A
NAME="GUC-LOG-HOSTNAME"
></A
><TT
CLASS="VARNAME"
>log_hostname</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> By default, connection log messages only show the IP address of the
connecting host. Turning on this option causes logging of the
host name as well. Note that depending on your host name resolution
setup this might impose a non-negligible performance penalty. This
option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> file.
</P
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-STATISTICS"
>16.4.7. Runtime Statistics</A
></H2
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-STATISTICS-MONITOR"
>16.4.7.1. Statistics Monitoring</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><TT
CLASS="VARNAME"
>log_statement_stats</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>log_parser_stats</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>log_planner_stats</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>log_executor_stats</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> For each query, write performance statistics of the respective
module to the server log. This is a crude profiling
instrument. <TT
CLASS="VARNAME"
>log_statement_stats</TT
> reports total
statement statistics, while the others report per-module statistics.
<TT
CLASS="VARNAME"
>log_statement_stats</TT
> cannot be enabled together with
any of the per-module options. All of these options are disabled by
default. Only superusers can change these settings.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-STATISTICS-COLLECTOR"
>16.4.7.2. Query and Index Statistics Collector</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-STATS-START-COLLECTOR"
></A
><TT
CLASS="VARNAME"
>stats_start_collector</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Controls whether the server should start the
statistics-collection subprocess. This is on by default, but
may be turned off if you know you have no interest in
collecting statistics. This option can only be set at server
start.
</P
></DD
><DT
><A
NAME="GUC-STATS-COMMAND-STRING"
></A
><TT
CLASS="VARNAME"
>stats_command_string</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables the collection of statistics on the currently
executing command of each session, along with the time at
which that command began execution. This option is off by
default. Note that even when enabled, this information is not
visible to all users, only to superusers and the user owning
the session being reported on; so it should not represent a
security risk. This data can be accessed via the
<TT
CLASS="STRUCTNAME"
>pg_stat_activity</TT
> system view; refer
to <A
HREF="monitoring.html"
>Chapter 23</A
> for more information.
</P
></DD
><DT
><A
NAME="GUC-STATS-BLOCK-LEVEL"
></A
><TT
CLASS="VARNAME"
>stats_block_level</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables the collection of block-level statistics on database
activity. This option is disabled by default. If this option
is enabled, the data that is produced can be accessed via the
<TT
CLASS="STRUCTNAME"
>pg_stat</TT
> and
<TT
CLASS="STRUCTNAME"
>pg_statio</TT
> family of system views;
refer to <A
HREF="monitoring.html"
>Chapter 23</A
> for more information.
</P
></DD
><DT
><A
NAME="GUC-STATS-ROW-LEVEL"
></A
><TT
CLASS="VARNAME"
>stats_row_level</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Enables the collection of row-level statistics on database
activity. This option is disabled by default. If this option
is enabled, the data that is produced can be accessed via the
<TT
CLASS="STRUCTNAME"
>pg_stat</TT
> and
<TT
CLASS="STRUCTNAME"
>pg_statio</TT
> family of system views;
refer to <A
HREF="monitoring.html"
>Chapter 23</A
> for more information.
</P
></DD
><DT
><A
NAME="GUC-STATS-RESET-ON-SERVER-START"
></A
><TT
CLASS="VARNAME"
>stats_reset_on_server_start</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> If on, collected statistics are zeroed out whenever the server
is restarted. If off, statistics are accumulated across server
restarts. The default is on. This option can only be set at
server start.
</P
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-CLIENT"
>16.4.8. Client Connection Defaults</A
></H2
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-CLIENT-STATEMENT"
>16.4.8.1. Statement Behavior</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-SEARCH-PATH"
></A
><TT
CLASS="VARNAME"
>search_path</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> This variable specifies the order in which schemas are searched
when an object (table, data type, function, etc.) is referenced by a
simple name with no schema component. When there are objects of
identical names in different schemas, the one found first
in the search path is used. An object that is not in any of the
schemas in the search path can only be referenced by specifying
its containing schema with a qualified (dotted) name.
</P
><P
> The value for <TT
CLASS="VARNAME"
>search_path</TT
> has to be a comma-separated
list of schema names. If one of the list items is
the special value <TT
CLASS="LITERAL"
>$user</TT
>, then the schema
having the name returned by <CODE
CLASS="FUNCTION"
>SESSION_USER</CODE
> is substituted, if there
is such a schema. (If not, <TT
CLASS="LITERAL"
>$user</TT
> is ignored.)
</P
><P
> The system catalog schema, <TT
CLASS="LITERAL"
>pg_catalog</TT
>, is always
searched, whether it is mentioned in the path or not. If it is
mentioned in the path then it will be searched in the specified
order. If <TT
CLASS="LITERAL"
>pg_catalog</TT
> is not in the path then it will
be searched <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>before</I
></SPAN
> searching any of the path items.
It should also be noted that the temporary-table schema,
<TT
CLASS="LITERAL"
>pg_temp_<TT
CLASS="REPLACEABLE"
><I
>nnn</I
></TT
></TT
>, is implicitly searched before any of
these.
</P
><P
> When objects are created without specifying a particular target
schema, they will be placed in the first schema listed
in the search path. An error is reported if the search path is
empty.
</P
><P
> The default value for this parameter is
<TT
CLASS="LITERAL"
>'$user, public'</TT
> (where the second part will be
ignored if there is no schema named <TT
CLASS="LITERAL"
>public</TT
>).
This supports shared use of a database (where no users
have private schemas, and all share use of <TT
CLASS="LITERAL"
>public</TT
>),
private per-user schemas, and combinations of these. Other
effects can be obtained by altering the default search path
setting, either globally or per-user.
</P
><P
> The current effective value of the search path can be examined
via the <ACRONYM
CLASS="ACRONYM"
>SQL</ACRONYM
> function
<CODE
CLASS="FUNCTION"
>current_schemas()</CODE
>. This is not quite the same as
examining the value of <TT
CLASS="VARNAME"
>search_path</TT
>, since
<CODE
CLASS="FUNCTION"
>current_schemas()</CODE
> shows how the requests
appearing in <TT
CLASS="VARNAME"
>search_path</TT
> were resolved.
</P
><P
> For more information on schema handling, see <A
HREF="ddl-schemas.html"
>Section 5.8</A
>.
</P
></DD
><DT
><A
NAME="GUC-DEFAULT-TABLESPACE"
></A
><TT
CLASS="VARNAME"
>default_tablespace</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> This variable specifies the default tablespace in which to create
objects (tables and indexes) when a <TT
CLASS="COMMAND"
>CREATE</TT
> command does
not explicitly specify a tablespace.
</P
><P
> The value is either the name of a tablespace, or an empty string
to specify using the default tablespace of the current database.
If the value does not match the name of any existing tablespace,
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> will automatically use the default
tablespace of the current database.
</P
><P
> For more information on tablespaces,
see <A
HREF="manage-ag-tablespaces.html"
>Section 18.6</A
>.
</P
></DD
><DT
><A
NAME="GUC-CHECK-FUNCTION-BODIES"
></A
><TT
CLASS="VARNAME"
>check_function_bodies</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> This parameter is normally true. When set to false, it disables
validation of the function body string during <A
HREF="sql-createfunction.html"
><I
>CREATE FUNCTION</I
></A
>. Disabling validation is
occasionally useful to avoid problems such as forward
references when restoring function definitions from a dump.
</P
></DD
><DT
><A
NAME="GUC-DEFAULT-TRANSACTION-ISOLATION"
></A
><TT
CLASS="VARNAME"
>default_transaction_isolation</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Each SQL transaction has an isolation level, which can be
either <SPAN
CLASS="QUOTE"
>"read uncommitted"</SPAN
>, <SPAN
CLASS="QUOTE"
>"read
committed"</SPAN
>, <SPAN
CLASS="QUOTE"
>"repeatable read"</SPAN
>, or
<SPAN
CLASS="QUOTE"
>"serializable"</SPAN
>. This parameter controls the
default isolation level of each new transaction. The default
is <SPAN
CLASS="QUOTE"
>"read committed"</SPAN
>.
</P
><P
> Consult <A
HREF="mvcc.html"
>Chapter 12</A
> and <A
HREF="sql-set-transaction.html"
>SET TRANSACTION</A
> for more
information.
</P
></DD
><DT
><A
NAME="GUC-DEFAULT-TRANSACTION-READ-ONLY"
></A
><TT
CLASS="VARNAME"
>default_transaction_read_only</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> A read-only SQL transaction cannot alter non-temporary tables.
This parameter controls the default read-only status of each new
transaction. The default is false (read/write).
</P
><P
> Consult <A
HREF="sql-set-transaction.html"
>SET TRANSACTION</A
> for more information.
</P
></DD
><DT
><A
NAME="GUC-STATEMENT-TIMEOUT"
></A
><TT
CLASS="VARNAME"
>statement_timeout</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Abort any statement that takes over the specified number of
milliseconds. A value of zero (the default) turns off the limitation.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-CLIENT-FORMAT"
>16.4.8.2. Locale and Formatting</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-DATESTYLE"
></A
><TT
CLASS="VARNAME"
>DateStyle</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Sets the display format for date and time values, as well as the
rules for interpreting ambiguous date input values. For
historical reasons, this variable contains two independent
components: the output format specification (<TT
CLASS="LITERAL"
>ISO</TT
>,
<TT
CLASS="LITERAL"
>Postgres</TT
>, <TT
CLASS="LITERAL"
>SQL</TT
>, or <TT
CLASS="LITERAL"
>German</TT
>)
and the input/output specification for year/month/day ordering
(<TT
CLASS="LITERAL"
>DMY</TT
>, <TT
CLASS="LITERAL"
>MDY</TT
>, or <TT
CLASS="LITERAL"
>YMD</TT
>). These
can be set separately or together. The keywords <TT
CLASS="LITERAL"
>Euro</TT
>
and <TT
CLASS="LITERAL"
>European</TT
> are synonyms for <TT
CLASS="LITERAL"
>DMY</TT
>; the
keywords <TT
CLASS="LITERAL"
>US</TT
>, <TT
CLASS="LITERAL"
>NonEuro</TT
>, and
<TT
CLASS="LITERAL"
>NonEuropean</TT
> are synonyms for <TT
CLASS="LITERAL"
>MDY</TT
>. See
<A
HREF="datatype-datetime.html"
>Section 8.5</A
> for more information. The
default is <TT
CLASS="LITERAL"
>ISO, MDY</TT
>.
</P
></DD
><DT
><A
NAME="GUC-TIMEZONE"
></A
><TT
CLASS="VARNAME"
>timezone</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Sets the time zone for displaying and interpreting time
stamps. The default is 'unknown', which means to use whatever
the system environment specifies as the time zone. See <A
HREF="datatype-datetime.html"
>Section 8.5</A
> for more information.
</P
></DD
><DT
><A
NAME="GUC-AUSTRALIAN-TIMEZONES"
></A
><TT
CLASS="VARNAME"
>australian_timezones</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> If set to true, <TT
CLASS="LITERAL"
>ACST</TT
>,
<TT
CLASS="LITERAL"
>CST</TT
>, <TT
CLASS="LITERAL"
>EST</TT
>, and
<TT
CLASS="LITERAL"
>SAT</TT
> are interpreted as Australian time
zones rather than as North/South American time zones and
Saturday. The default is false.
</P
></DD
><DT
><A
NAME="GUC-EXTRA-FLOAT-DIGITS"
></A
><TT
CLASS="VARNAME"
>extra_float_digits</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> This parameter adjusts the number of digits displayed for
floating-point values, including <TT
CLASS="TYPE"
>float4</TT
>, <TT
CLASS="TYPE"
>float8</TT
>,
and geometric data types. The parameter value is added to the
standard number of digits (<TT
CLASS="LITERAL"
>FLT_DIG</TT
> or <TT
CLASS="LITERAL"
>DBL_DIG</TT
>
as appropriate). The value can be set as high as 2, to include
partially-significant digits; this is especially useful for dumping
float data that needs to be restored exactly. Or it can be set
negative to suppress unwanted digits.
</P
></DD
><DT
><A
NAME="GUC-CLIENT-ENCODING"
></A
><TT
CLASS="VARNAME"
>client_encoding</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Sets the client-side encoding (character set).
The default is to use the database encoding.
</P
></DD
><DT
><A
NAME="GUC-LC-MESSAGES"
></A
><TT
CLASS="VARNAME"
>lc_messages</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Sets the language in which messages are displayed. Acceptable
values are system-dependent; see <A
HREF="charset.html#LOCALE"
>Section 20.1</A
> for
more information. If this variable is set to the empty string
(which is the default) then the value is inherited from the
execution environment of the server in a system-dependent way.
</P
><P
> On some systems, this locale category does not exist. Setting
this variable will still work, but there will be no effect.
Also, there is a chance that no translated messages for the
desired language exist. In that case you will continue to see
the English messages.
</P
></DD
><DT
><A
NAME="GUC-LC-MONETARY"
></A
><TT
CLASS="VARNAME"
>lc_monetary</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Sets the locale to use for formatting monetary amounts, for
example with the <CODE
CLASS="FUNCTION"
>to_char</CODE
> family of
functions. Acceptable values are system-dependent; see <A
HREF="charset.html#LOCALE"
>Section 20.1</A
> for more information. If this variable is
set to the empty string (which is the default) then the value
is inherited from the execution environment of the server in a
system-dependent way.
</P
></DD
><DT
><A
NAME="GUC-LC-NUMERIC"
></A
><TT
CLASS="VARNAME"
>lc_numeric</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Sets the locale to use for formatting numbers, for example
with the <CODE
CLASS="FUNCTION"
>to_char</CODE
> family of
functions. Acceptable values are system-dependent; see <A
HREF="charset.html#LOCALE"
>Section 20.1</A
> for more information. If this variable is
set to the empty string (which is the default) then the value
is inherited from the execution environment of the server in a
system-dependent way.
</P
></DD
><DT
><A
NAME="GUC-LC-TIME"
></A
><TT
CLASS="VARNAME"
>lc_time</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Sets the locale to use for formatting date and time values.
(Currently, this setting does nothing, but it may in the
future.) Acceptable values are system-dependent; see <A
HREF="charset.html#LOCALE"
>Section 20.1</A
> for more information. If this variable is
set to the empty string (which is the default) then the value
is inherited from the execution environment of the server in a
system-dependent way.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-CLIENT-OTHER"
>16.4.8.3. Other Defaults</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-EXPLAIN-PRETTY-PRINT"
></A
><TT
CLASS="VARNAME"
>explain_pretty_print</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Determines whether <TT
CLASS="COMMAND"
>EXPLAIN VERBOSE</TT
> uses the
indented or non-indented format for displaying detailed
query-tree dumps. The default is on.
</P
></DD
><DT
><A
NAME="GUC-DYNAMIC-LIBRARY-PATH"
></A
><TT
CLASS="VARNAME"
>dynamic_library_path</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> If a dynamically loadable module needs to be opened and the
file name specified in the <TT
CLASS="COMMAND"
>CREATE FUNCTION</TT
> or
<TT
CLASS="COMMAND"
>LOAD</TT
> command
does not have a directory component (i.e. the
name does not contain a slash), the system will search this
path for the required file.
</P
><P
> The value for <TT
CLASS="VARNAME"
>dynamic_library_path</TT
> has to be a
list of absolute directory paths separated by colons (or semi-colons
on Windows). If a list element starts
with the special string <TT
CLASS="LITERAL"
>$libdir</TT
>, the
compiled-in <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> package
library directory is substituted for <TT
CLASS="LITERAL"
>$libdir</TT
>. This
is where the modules provided by the standard
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> distribution are installed.
(Use <TT
CLASS="LITERAL"
>pg_config --pkglibdir</TT
> to find out the name of
this directory.) For example:
</P><PRE
CLASS="PROGRAMLISTING"
>dynamic_library_path = '/usr/local/lib/postgresql:/home/my_project/lib:$libdir'</PRE
><P>
or, in a Windows environment:
</P><PRE
CLASS="PROGRAMLISTING"
>dynamic_library_path = 'C:\tools\postgresql;H:\my_project\lib;$libdir'</PRE
><P>
</P
><P
> The default value for this parameter is
<TT
CLASS="LITERAL"
>'$libdir'</TT
>. If the value is set to an empty
string, the automatic path search is turned off.
</P
><P
> This parameter can be changed at run time by superusers, but a
setting done that way will only persist until the end of the
client connection, so this method should be reserved for
development purposes. The recommended way to set this parameter
is in the <TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration
file.
</P
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-LOCKS"
>16.4.9. Lock Management</A
></H2
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-DEADLOCK-TIMEOUT"
></A
><TT
CLASS="VARNAME"
>deadlock_timeout</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> This is the amount of time, in milliseconds, to wait on a lock
before checking to see if there is a deadlock condition. The
check for deadlock is relatively slow, so the server doesn't run
it every time it waits for a lock. We (optimistically?) assume
that deadlocks are not common in production applications and
just wait on the lock for a while before starting the check for a
deadlock. Increasing this value reduces the amount of time
wasted in needless deadlock checks, but slows down reporting of
real deadlock errors. The default is 1000 (i.e., one second),
which is probably about the smallest value you would want in
practice. On a heavily loaded server you might want to raise it.
Ideally the setting should exceed your typical transaction time,
so as to improve the odds that a lock will be released before
the waiter decides to check for deadlock.
</P
></DD
><DT
><A
NAME="GUC-MAX-LOCKS-PER-TRANSACTION"
></A
><TT
CLASS="VARNAME"
>max_locks_per_transaction</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> The shared lock table is sized on the assumption that at most
<TT
CLASS="VARNAME"
>max_locks_per_transaction</TT
> *
<TT
CLASS="VARNAME"
>max_connections</TT
> distinct objects will need to
be locked at any one time. (Thus, this parameter's name may be
confusing: it is not a hard limit on the number of locks taken
by any one transaction, but rather a maximum average value.)
The default, 64, has historically
proven sufficient, but you might need to raise this value if you
have clients that touch many different tables in a single
transaction. This option can only be set at server start.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-COMPATIBLE"
>16.4.10. Version and Platform Compatibility</A
></H2
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-COMPATIBLE-VERSION"
>16.4.10.1. Previous PostgreSQL Versions</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-ADD-MISSING-FROM"
></A
><TT
CLASS="VARNAME"
>add_missing_from</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> When <TT
CLASS="LITERAL"
>true</TT
>, tables that are referenced by a query will be
automatically added to the <TT
CLASS="LITERAL"
>FROM</TT
> clause if not already
present. The default is <TT
CLASS="LITERAL"
>true</TT
> for compatibility with
previous releases of <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>. However, this
behavior is not SQL-standard, and many people dislike it because it
can mask mistakes (such as referencing a table where you should have
referenced its alias). Set to <TT
CLASS="LITERAL"
>false</TT
> for the SQL-standard
behavior of rejecting references to tables that are not listed in
<TT
CLASS="LITERAL"
>FROM</TT
>.
</P
></DD
><DT
><A
NAME="GUC-BACKSLASH-QUOTE"
></A
><TT
CLASS="VARNAME"
>backslash_quote</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> This controls whether a quote mark can be represented by
<TT
CLASS="LITERAL"
>\'</TT
> in a string literal. The preferred, SQL-standard way
to represent a quote mark is by doubling it (<TT
CLASS="LITERAL"
>''</TT
>) but
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> has historically also accepted
<TT
CLASS="LITERAL"
>\'</TT
>. However, use of <TT
CLASS="LITERAL"
>\'</TT
> creates security risks
because in some client character set encodings, there are multibyte
characters in which the last byte is numerically equivalent to ASCII
<TT
CLASS="LITERAL"
>\</TT
>. If client-side code does escaping incorrectly then a
SQL-injection attack is possible. This risk can be prevented by
making the server reject queries in which a quote mark appears to be
escaped by a backslash.
The allowed values of <TT
CLASS="VARNAME"
>backslash_quote</TT
> are
<TT
CLASS="LITERAL"
>on</TT
> (allow <TT
CLASS="LITERAL"
>\'</TT
> always),
<TT
CLASS="LITERAL"
>off</TT
> (reject always), and
<TT
CLASS="LITERAL"
>safe_encoding</TT
> (allow only if client encoding does not
allow ASCII <TT
CLASS="LITERAL"
>\</TT
> within a multibyte character).
<TT
CLASS="LITERAL"
>safe_encoding</TT
> is the default setting.
</P
></DD
><DT
><A
NAME="GUC-DEFAULT-WITH-OIDS"
></A
><TT
CLASS="VARNAME"
>default_with_oids</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> This controls whether <TT
CLASS="COMMAND"
>CREATE TABLE</TT
> and
<TT
CLASS="COMMAND"
>CREATE TABLE AS</TT
> include an OID column in
newly-created tables, if neither <TT
CLASS="LITERAL"
>WITH OIDS</TT
>
nor <TT
CLASS="LITERAL"
>WITHOUT OIDS</TT
> is specified. It also
determines whether OIDs will be included in tables created by
<TT
CLASS="COMMAND"
>SELECT INTO</TT
>. In
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> 8.0.11
<TT
CLASS="VARNAME"
>default_with_oids</TT
> defaults to true. This is
also the behavior of previous versions of
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>. However, assuming that
tables will contain OIDs by default is not encouraged. This
option will probably default to false in a future release of
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>.
</P
><P
> To ease compatibility with applications that make use of OIDs,
this option should left enabled. To ease compatibility with
future versions of <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>, this
option should be disabled, and applications that require OIDs
on certain tables should explicitly specify <TT
CLASS="LITERAL"
>WITH
OIDS</TT
> when those tables are created.
</P
></DD
><DT
><A
NAME="GUC-REGEX-FLAVOR"
></A
><TT
CLASS="VARNAME"
>regex_flavor</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> The regular expression <SPAN
CLASS="QUOTE"
>"flavor"</SPAN
> can be set to
<TT
CLASS="LITERAL"
>advanced</TT
>, <TT
CLASS="LITERAL"
>extended</TT
>, or <TT
CLASS="LITERAL"
>basic</TT
>.
The default is <TT
CLASS="LITERAL"
>advanced</TT
>. The <TT
CLASS="LITERAL"
>extended</TT
>
setting may be useful for exact backwards compatibility with
pre-7.4 releases of <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>. See
<A
HREF="functions-matching.html#POSIX-SYNTAX-DETAILS"
>Section 9.7.3.1</A
> for details.
</P
></DD
><DT
><A
NAME="GUC-SQL-INHERITANCE"
></A
><TT
CLASS="VARNAME"
>sql_inheritance</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> This controls the inheritance semantics, in particular whether
subtables are included by various commands by default. They were
not included in versions prior to 7.1. If you need the old
behavior you can set this variable to off, but in the long run
you are encouraged to change your applications to use the
<TT
CLASS="LITERAL"
>ONLY</TT
> key word to exclude subtables. See
<A
HREF="ddl-inherit.html"
>Section 5.5</A
> for more information about inheritance.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="RUNTIME-CONFIG-COMPATIBLE-CLIENTS"
>16.4.10.2. Platform and Client Compatibility</A
></H3
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-TRANSFORM-NULL-EQUALS"
></A
><TT
CLASS="VARNAME"
>transform_null_equals</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> When turned on, expressions of the form
<TT
CLASS="LITERAL"
><TT
CLASS="REPLACEABLE"
><I
>expr</I
></TT
> = NULL</TT
> (or <TT
CLASS="LITERAL"
>NULL
= <TT
CLASS="REPLACEABLE"
><I
>expr</I
></TT
></TT
>) are treated as
<TT
CLASS="LITERAL"
><TT
CLASS="REPLACEABLE"
><I
>expr</I
></TT
> IS NULL</TT
>, that is, they
return true if <TT
CLASS="REPLACEABLE"
><I
>expr</I
></TT
> evaluates to the null value,
and false otherwise. The correct SQL-spec-compliant behavior of
<TT
CLASS="LITERAL"
><TT
CLASS="REPLACEABLE"
><I
>expr</I
></TT
> = NULL</TT
> is to always
return null (unknown). Therefore this option defaults to off.
</P
><P
> However, filtered forms in <SPAN
CLASS="PRODUCTNAME"
>Microsoft
Access</SPAN
> generate queries that appear to use
<TT
CLASS="LITERAL"
><TT
CLASS="REPLACEABLE"
><I
>expr</I
></TT
> = NULL</TT
> to test for
null values, so if you use that interface to access the database you
might want to turn this option on. Since expressions of the
form <TT
CLASS="LITERAL"
><TT
CLASS="REPLACEABLE"
><I
>expr</I
></TT
> = NULL</TT
> always
return the null value (using the correct interpretation) they are not
very useful and do not appear often in normal applications, so
this option does little harm in practice. But new users are
frequently confused about the semantics of expressions
involving null values, so this option is not on by default.
</P
><P
> Note that this option only affects the exact form <TT
CLASS="LITERAL"
>= NULL</TT
>,
not other comparison operators or other expressions
that are computationally equivalent to some expression
involving the equals operator (such as <TT
CLASS="LITERAL"
>IN</TT
>).
Thus, this option is not a general fix for bad programming.
</P
><P
> Refer to <A
HREF="functions-comparison.html"
>Section 9.2</A
> for related information.
</P
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-PRESET"
>16.4.11. Preset Options</A
></H2
><P
> The following <SPAN
CLASS="QUOTE"
>"parameters"</SPAN
> are read-only, and are determined
when <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> is compiled or when it is
installed. As such, they have been excluded from the sample
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> file. These options report
various aspects of <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> behavior
that may be of interest to certain applications, particularly
administrative front-ends.
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-BLOCK-SIZE"
></A
><TT
CLASS="VARNAME"
>block_size</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Shows the size of a disk block. It is determined by the value
of <TT
CLASS="LITERAL"
>BLCKSZ</TT
> when building the server. The default
value is 8192 bytes. The meaning of some configuration
variables (such as <A
HREF="runtime-config.html#GUC-SHARED-BUFFERS"
>shared_buffers</A
>) is
influenced by <TT
CLASS="VARNAME"
>block_size</TT
>. See <A
HREF="runtime-config.html#RUNTIME-CONFIG-RESOURCE"
>Section 16.4.3</A
> for information.
</P
></DD
><DT
><A
NAME="GUC-INTEGER-DATETIMES"
></A
><TT
CLASS="VARNAME"
>integer_datetimes</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Shows whether <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> was built
with support for 64-bit-integer dates and times. It is set by
configuring with <TT
CLASS="LITERAL"
>--enable-integer-datetimes</TT
>
when building <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>. The
default value is <TT
CLASS="LITERAL"
>off</TT
>.
</P
></DD
><DT
><A
NAME="GUC-LC-COLLATE"
></A
><TT
CLASS="VARNAME"
>lc_collate</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Shows the locale in which sorting of textual data is done.
See <A
HREF="charset.html#LOCALE"
>Section 20.1</A
> for more information.
The value is determined when the database cluster is initialized.
</P
></DD
><DT
><A
NAME="GUC-LC-CTYPE"
></A
><TT
CLASS="VARNAME"
>lc_ctype</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Shows the locale that determines character classifications.
See <A
HREF="charset.html#LOCALE"
>Section 20.1</A
> for more information.
The value is determined when the database cluster is initialized.
Ordinarily this will be the same as <TT
CLASS="VARNAME"
>lc_collate</TT
>,
but for special applications it might be set differently.
</P
></DD
><DT
><A
NAME="GUC-MAX-FUNCTION-ARGS"
></A
><TT
CLASS="VARNAME"
>max_function_args</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Shows the maximum number of function arguments. It is determined by
the value of <TT
CLASS="LITERAL"
>FUNC_MAX_ARGS</TT
> when building the server. The
default value is 32.
</P
></DD
><DT
><A
NAME="GUC-MAX-IDENTIFIER-LENGTH"
></A
><TT
CLASS="VARNAME"
>max_identifier_length</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Shows the maximum identifier length. It is determined as one
less than the value of <TT
CLASS="LITERAL"
>NAMEDATALEN</TT
> when building
the server. The default value of <TT
CLASS="LITERAL"
>NAMEDATALEN</TT
> is
64; therefore the default
<TT
CLASS="VARNAME"
>max_identifier_length</TT
> is 63.
</P
></DD
><DT
><A
NAME="GUC-MAX-INDEX-KEYS"
></A
><TT
CLASS="VARNAME"
>max_index_keys</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Shows the maximum number of index keys. It is determined by
the value of <TT
CLASS="LITERAL"
>INDEX_MAX_KEYS</TT
> when building the server. The
default value is 32.
</P
></DD
><DT
><A
NAME="GUC-SERVER-ENCODING"
></A
><TT
CLASS="VARNAME"
>server_encoding</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Shows the database encoding (character set).
It is determined when the database is created. Ordinarily,
clients need only be concerned with the value of <A
HREF="runtime-config.html#GUC-CLIENT-ENCODING"
>client_encoding</A
>.
</P
></DD
><DT
><A
NAME="GUC-SERVER-VERSION"
></A
><TT
CLASS="VARNAME"
>server_version</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> Shows the version number of the server. It is determined by the
value of <TT
CLASS="LITERAL"
>PG_VERSION</TT
> when building the server.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-CUSTOM"
>16.4.12. Customized Options</A
></H2
><P
> This feature was designed to allow options not normally known to
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> to be added by add-on modules
(such as procedural languages). This allows add-on modules to be
configured in the standard ways.
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-CUSTOM-VARIABLE-CLASSES"
></A
><TT
CLASS="VARNAME"
>custom_variable_classes</TT
> (<TT
CLASS="TYPE"
>string</TT
>)</DT
><DD
><P
> This variable specifies one or several class names to be used for
custom variables, in the form of a comma-separated list. A custom
variable is a variable not normally known
to <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> proper but used by some
add-on module. Such variables must have names consisting of a class
name, a dot, and a variable name. <TT
CLASS="VARNAME"
>custom_variable_classes</TT
>
specifies all the class names in use in a particular installation.
This option can only be set at server start or in the
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> configuration file.
</P
></DD
></DL
></DIV
><P
> The difficulty with setting custom variables in
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> is that the file must be read before add-on
modules have been loaded, and so custom variables would ordinarily be
rejected as unknown. When <TT
CLASS="VARNAME"
>custom_variable_classes</TT
> is set,
the server will accept definitions of arbitrary variables within each
specified class. These variables will be treated as placeholders and
will have no function until the module that defines them is loaded. When a
module for a specific class is loaded, it will add the proper variable
definitions for its class name, convert any placeholder
values according to those definitions, and issue warnings for any
placeholders of its class that remain (which presumably would be
misspelled configuration variables).
</P
><P
> Here is an example of what <TT
CLASS="FILENAME"
>postgresql.conf</TT
> might contain
when using custom variables:
</P><PRE
CLASS="PROGRAMLISTING"
>custom_variable_classes = 'plr,pljava'
plr.path = '/usr/lib/R'
pljava.foo = 1
plruby.bar = true # generates error, unknown class name</PRE
><P>
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-DEVELOPER"
>16.4.13. Developer Options</A
></H2
><P
> The following options are intended for work on the
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> source, and in some cases
to assist with recovery of severely damaged databases. There
should be no reason to use them in a production database setup.
As such, they have been excluded from the sample
<TT
CLASS="FILENAME"
>postgresql.conf</TT
> file. Note that many of these
options require special source compilation flags to work at all.
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="GUC-DEBUG-ASSERTIONS"
></A
><TT
CLASS="VARNAME"
>debug_assertions</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Turns on various assertion checks. This is a debugging aid. If
you are experiencing strange problems or crashes you might want
to turn this on, as it might expose programming mistakes. To use
this option, the macro <TT
CLASS="SYMBOL"
>USE_ASSERT_CHECKING</TT
>
must be defined when <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> is
built (accomplished by the <TT
CLASS="COMMAND"
>configure</TT
> option
<TT
CLASS="OPTION"
>--enable-cassert</TT
>). Note that
<TT
CLASS="VARNAME"
>debug_assertions</TT
> defaults to on if
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> has been built with
assertions enabled.
</P
></DD
><DT
><A
NAME="GUC-DEBUG-SHARED-BUFFERS"
></A
><TT
CLASS="VARNAME"
>debug_shared_buffers</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> Number of seconds between buffer freelist reports.
If set greater than zero, emit freelist statistics to the log every so many
seconds. Zero (the default) disables reporting.
</P
></DD
><DT
><A
NAME="GUC-PRE-AUTH-DELAY"
></A
><TT
CLASS="VARNAME"
>pre_auth_delay</TT
> (<TT
CLASS="TYPE"
>integer</TT
>)</DT
><DD
><P
> If nonzero, a delay of this many seconds occurs just after a new
server process is forked, before it conducts the authentication
process. This is intended to give an opportunity to attach to the
server process with a debugger to trace down misbehavior in
authentication.
</P
></DD
><DT
><A
NAME="GUC-TRACE-NOTIFY"
></A
><TT
CLASS="VARNAME"
>trace_notify</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Generates a great amount of debugging output for the
<TT
CLASS="COMMAND"
>LISTEN</TT
> and <TT
CLASS="COMMAND"
>NOTIFY</TT
>
commands. <A
HREF="runtime-config.html#GUC-CLIENT-MIN-MESSAGES"
>client_min_messages</A
> or
<A
HREF="runtime-config.html#GUC-LOG-MIN-MESSAGES"
>log_min_messages</A
> must be
<TT
CLASS="LITERAL"
>DEBUG1</TT
> or lower to send this output to the
client or server log, respectively.
</P
></DD
><DT
><TT
CLASS="VARNAME"
>trace_locks</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>trace_lwlocks</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>trace_userlocks</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>trace_lock_oidmin</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>trace_lock_table</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>debug_deadlocks</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)<BR><TT
CLASS="VARNAME"
>log_btree_build_stats</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Various other code tracing and debugging options.
</P
></DD
><DT
><A
NAME="GUC-WAL-DEBUG"
></A
><TT
CLASS="VARNAME"
>wal_debug</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> If true, emit WAL-related debugging output. This option is
only available if the <TT
CLASS="SYMBOL"
>WAL_DEBUG</TT
> macro was
defined when <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> was
compiled.
</P
></DD
><DT
><A
NAME="GUC-ZERO-DAMAGED-PAGES"
></A
><TT
CLASS="VARNAME"
>zero_damaged_pages</TT
> (<TT
CLASS="TYPE"
>boolean</TT
>)</DT
><DD
><P
> Detection of a damaged page header normally causes
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> to report an error, aborting the current
command. Setting <TT
CLASS="VARNAME"
>zero_damaged_pages</TT
> to true causes
the system to instead report a warning, zero out the damaged page,
and continue processing. This behavior <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>will destroy data</I
></SPAN
>,
namely all the rows on the damaged page. But it allows you to get
past the error and retrieve rows from any undamaged pages that may
be present in the table. So it is useful for recovering data if
corruption has occurred due to hardware or software error. You should
generally not set this true until you have given up hope of recovering
data from the damaged page(s) of a table. The
default setting is off, and it can only be changed by a superuser.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="RUNTIME-CONFIG-SHORT"
>16.4.14. Short Options</A
></H2
><P
> For convenience there are also single letter command-line option switches
available for some parameters. They are described in <A
HREF="runtime-config.html#RUNTIME-CONFIG-SHORT-TABLE"
>Table 16-1</A
>.
</P
><DIV
CLASS="TABLE"
><A
NAME="RUNTIME-CONFIG-SHORT-TABLE"
></A
><P
><B
>Table 16-1. Short option key</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><THEAD
><TR
><TH
>Short option</TH
><TH
>Equivalent</TH
></TR
></THEAD
><TBODY
><TR
><TD
><TT
CLASS="OPTION"
>-B <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
><TD
><TT
CLASS="LITERAL"
>shared_buffers = <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-d <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
><TD
><TT
CLASS="LITERAL"
>log_min_messages = DEBUG<TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-F</TT
></TD
><TD
><TT
CLASS="LITERAL"
>fsync = off</TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-h <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
><TD
><TT
CLASS="LITERAL"
>listen_addresses = <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-i</TT
></TD
><TD
><TT
CLASS="LITERAL"
>listen_addresses = '*'</TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-k <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
><TD
><TT
CLASS="LITERAL"
>unix_socket_directory = <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-l</TT
></TD
><TD
><TT
CLASS="LITERAL"
>ssl = on</TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-N <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
><TD
><TT
CLASS="LITERAL"
>max_connections = <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-p <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
><TD
><TT
CLASS="LITERAL"
>port = <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
></TR
><TR
><TD
> <TT
CLASS="OPTION"
>-fi</TT
>, <TT
CLASS="OPTION"
>-fh</TT
>,
<TT
CLASS="OPTION"
>-fm</TT
>, <TT
CLASS="OPTION"
>-fn</TT
>,
<TT
CLASS="OPTION"
>-fs</TT
>, <TT
CLASS="OPTION"
>-ft</TT
><A
NAME="FN.RUNTIME-CONFIG-SHORT"
HREF="#FTN.FN.RUNTIME-CONFIG-SHORT"
><SPAN
CLASS="footnote"
>[a]</SPAN
></A
>
</TD
><TD
> <TT
CLASS="LITERAL"
>enable_indexscan = off</TT
>,
<TT
CLASS="LITERAL"
>enable_hashjoin = off</TT
>,
<TT
CLASS="LITERAL"
>enable_mergejoin = off</TT
>,
<TT
CLASS="LITERAL"
>enable_nestloop = off</TT
>,
<TT
CLASS="LITERAL"
>enable_seqscan = off</TT
>,
<TT
CLASS="LITERAL"
>enable_tidscan = off</TT
>
</TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-s</TT
><A
HREF="runtime-config.html#FTN.FN.RUNTIME-CONFIG-SHORT"
><SPAN
CLASS="footnote"
>[a]</SPAN
></A
></TD
><TD
><TT
CLASS="LITERAL"
>log_statement_stats = on</TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-S <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
><A
HREF="runtime-config.html#FTN.FN.RUNTIME-CONFIG-SHORT"
><SPAN
CLASS="footnote"
>[a]</SPAN
></A
>
</TD
><TD
><TT
CLASS="LITERAL"
>work_mem = <TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
></TT
></TD
></TR
><TR
><TD
><TT
CLASS="OPTION"
>-tpa</TT
>, <TT
CLASS="OPTION"
>-tpl</TT
>, <TT
CLASS="OPTION"
>-te</TT
><A
HREF="runtime-config.html#FTN.FN.RUNTIME-CONFIG-SHORT"
><SPAN
CLASS="footnote"
>[a]</SPAN
></A
></TD
><TD
><TT
CLASS="LITERAL"
>log_parser_stats = on</TT
>,
<TT
CLASS="LITERAL"
>log_planner_stats = on</TT
>,
<TT
CLASS="LITERAL"
>log_executor_stats = on</TT
></TD
></TR
></TBODY
><TR
><TD
COLSPAN="2"
>Notes:<BR><A
NAME="FTN.FN.RUNTIME-CONFIG-SHORT"
>a. </A
> For historical reasons, these options must be passed to
the individual server process via the <TT
CLASS="OPTION"
>-o</TT
>
<TT
CLASS="COMMAND"
>postmaster</TT
> option, for example,
</P><PRE
CLASS="SCREEN"
>$ <KBD
CLASS="USERINPUT"
>postmaster -o '-S 1024 -s'</KBD
></PRE
><P>
or via <TT
CLASS="ENVAR"
>PGOPTIONS</TT
> from the client side, as
explained above.
<BR></TD
></TR
></TABLE
></DIV
></DIV
></DIV
><DIV
CLASS="NAVFOOTER"
><HR
ALIGN="LEFT"
WIDTH="100%"><TABLE
SUMMARY="Footer navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
><A
HREF="postmaster-start.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="index.html"
ACCESSKEY="H"
>Home</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
><A
HREF="kernel-resources.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
>Starting the Database Server</TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="runtime.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
>Managing Kernel Resources</TD
></TR
></TABLE
></DIV
></BODY
></HTML
>
