<!--

$Header: /cvsroot/pgsql/doc/src/sgml/runtime.sgml,v 1.220 2003/11/1603:58:10 tgl Exp $

$Header: /cvsroot/pgsql/doc/src/sgml/runtime.sgml,v 1.221 2003/11/1617:01:50 tgl Exp $

-->

<Chapter Id="runtime">

<title>Linux Memory Overcommit</title>

<para>

In Linux 2.4, the default virtual memory behavior is not

In Linux 2.4 and later, the default virtual memory behavior is not

optimal for <productname>PostgreSQL</productname>. Because of the

way that the kernel implements memory overcommit, the kernel may

terminate the <productname>PostgreSQL</productname> server (the

Out of Memory: Killed process 12345 (postmaster).

</programlisting>

This indicates that the <filename>postmaster</filename> process

has been terminated due to memory pressure;

has been terminated due to memory pressure.

Although existing database connections will continue to function

normally, no new connections will be accepted. To recover,

<productname>PostgreSQL</productname> will need to be restarted.

</para>

<para>

To avoid this situation, run <productname>PostgreSQL</productname>

One way to avoid this problem is to run

<productname>PostgreSQL</productname>

on a machine where you can be sure that other processes will not

run the machine out of memory. If your kernel supports the strict

determined by examining the function <function>vm_enough_memory</>

file reveals thatthe strict or paranoid modes are supported by

run the machine out of memory.

<para>

behavior so thatit will not <quote>overcommit</> memory. This is

<programlisting>

sysctl -w vm.overcommit_memory=2

</programlisting>

for strict mode, or

<programlisting>

sysctl -w vm.overcommit_memory=3

</programlisting>

for paranoid mode, or placing an equivalent entry in

<filename>/etc/sysctl.conf</>. For more information on memory

overcommit handling and the meaning of these kernel settings,

consult the file

<filename>Documentation/vm/overcommit-accounting</filename> in a

recent version of the source code for the Linux kernel.

or placing an equivalent entry in <filename>/etc/sysctl.conf</>.

You may also wish to modify the related setting

<literal>vm.overcommit_ratio</>. For details see the kernel documentation

file <filename>Documentation/vm/overcommit-accounting</>.

</para>

<caution>

<para>

Using these settings with a kernel which does not support these

modes will almost certainly increase the danger of the kernel

killing the database server, rather than reducing it. If in any

doubt, consult a kernel expert or your kernel vendor.

</para>

</caution>

<para>

The method described above for changing the kernel's overcommit

behavior should work with Linux kernel versions 2.6 and later. In addition,

some versions of the 2.4 kernel distributed by some vendors may

also support this functionality. However, it is known that some

vendor documents suggest that they support them while examination

of the kernel source reveals that they do not.

Some vendors' Linux 2.4 kernels are reported to have early versions

of the 2.6 overcommit sysctl. However, setting

<literal>vm.overcommit_memory</> to 2

on a kernel that does not have the relevant code will make

things worse not better. It is recommended that you inspect

the actual kernel source code (see the function

<function>vm_enough_memory</> in the file <filename>mm/mmap.c</>)

to verify what is supported in your copy before you try this in a 2.4

installation. The presence of the <filename>overcommit-accounting</>

documentation file should <emphasis>not</> be taken as evidence that the

feature is there. If in any doubt, consult a kernel expert or your

kernel vendor.

</para>

</sect2>

</sect1>