<?xml version="1.0"?>

<!DOCTYPE module SYSTEM "../../../../dtd/module.dtd">

<module name="Module ngx_http_upstream_module"
        link="/en/docs/http/ngx_http_upstream_module.html"
        lang="en">

<section id="summary">

<para>
The <literal>ngx_http_upstream_module</literal> module
allows to define groups of servers that can be referenced
from the <link doc="ngx_http_proxy_module.xml" id="proxy_pass"/> and
<link doc="ngx_http_fastcgi_module.xml" id="fastcgi_pass"/> directives.
</para>

</section>


<section id="example" name="Example Configuration">

<para>
<example>
upstream <emphasis>backend</emphasis> {
    server backend1.example.com       weight=5;
    server backend2.example.com:8080;
    server unix:/tmp/backend3;

    server backup1.example.com:8080   backup;
    server backup2.example.com:8080   backup;
}

server {
    location / {
        proxy_pass http://<emphasis>backend</emphasis>;
    }
}
</example>
</para>

</section>


<section id="directives" name="Directives">

<directive name="ip_hash">
<syntax/>
<default/>
<context>upstream</context>

<para>
Specifies that a group should use a balancing method where requests
are distributed between servers based on client IP addresses.
A class C network containing the client IP address is used as a hashing key.
The method ensures that requests of the same client will always be
passed to the same server except when this server is considered down
in which case client requests will be passed to another server and
most probably it will also be the same server.
</para>

<para>
It is not possible to specify a weight for servers using the
<literal>ip_hash</literal> balancing method.
If one of the servers needs to be temporarily removed, it should
be marked with the <literal>down</literal> parameter in
order to preserve the current hashing of client IP addresses.
</para>

<para>
Example:
<example>
upstream backend {
    ip_hash;

    server backend1.example.com;
    server backend2.example.com;
    server backend3.example.com <emphasis>down</emphasis>;
    server backend4.example.com;
}
</example>
</para>

</directive>


<directive name="keepalive">
<syntax><value>connections</value></syntax>
<default/>
<context>upstream</context>
<appeared-in>1.1.4</appeared-in>

<para>
Activates cache of connections to upstream servers.
</para>

<para>
The <value>connections</value> parameter sets the maximum number of
keepalive connections to upstream servers that are retained in
the cache per one worker process.
If there are more idle connections to retain to an upstream than
are specified by this number, the least recently used ones will
be closed.
<note>
It should be noted that <literal>keepalive</literal> directive
does not limit the total number of connections that nginx worker
process can open to upstream servers.
The new connections are always created on demand.
The <value>connections</value> parameter should be set low enough
to allow upstream servers to process additional new incoming
connections as well.
</note>
</para>

<para>
Example configuration of memcached upstream with keepalive connections:
<example>
upstream memcached_backend {
    server 127.0.0.1:11211;
    server 10.0.0.2:11211;

    keepalive 32;
}

server {
    ...

    location /memcached/ {
        set $memcached_key $uri;
        memcached_pass memcached_backend;
    }

}
</example>
</para>

<para>
For HTTP, the <link doc="ngx_http_proxy_module.xml" id="proxy_http_version"/>
directive should be set to “<literal>1.1</literal>”
and the <header>Connection</header> header field should be cleared:
<example>
upstream http_backend {
    server 127.0.0.1:8080;

    keepalive 16;
}

server {
    ...

    location /http/ {
        proxy_pass http://http_backend;
        proxy_http_version 1.1;
        proxy_set_header Connection "";
        ...
    }
}
</example>
</para>

<para>
<note>
Alternatively, HTTP/1.0 persistent connections can be used by passing the
<header>Connection: Keep-Alive</header> header field to an upstream server,
though this is not recommended.
</note>
</para>

<para>
For FastCGI servers, it is required to set
<link doc="ngx_http_fastcgi_module.xml" id="fastcgi_keep_conn"/>
for keepalive connections to work:
<example>
upstream fastcgi_backend {
    server 127.0.0.1:9000;

    keepalive 8;
}

server {
    ...

    location /fastcgi/ {
        fastcgi_pass fastcgi_backend;
        fastcgi_keep_conn on;
        ...
    }
}
</example>
</para>

<para>
<note>
When using load balancer modules other than the default
round-robin, it is necessary to activate them before
the <literal>keepalive</literal> directive.
</note>

<note>
SCGI and uwsgi protocols do not define keepalive connections.
</note>
</para>

</directive>


<directive name="server">
<syntax><value>name</value> [<value>parameters</value>]</syntax>
<default/>
<context>upstream</context>

<para>
Sets a <value>name</value> and other <value>parameters</value> of the server.
A <value>name</value> can be a domain name, an address, a port, or a
path of the UNIX-domain socket.
If a domain name resolves to several addresses, all of them are used.
<list type="tag">

<tag-name><literal>weight</literal>=<value>number</value></tag-name>
<tag-desc>
sets a weight of the server, by default 1.
</tag-desc>

<tag-name><literal>max_fails</literal>=<value>number</value></tag-name>
<tag-desc>
sets a number of unsuccessful attempts to communicate with the server
during a time set by the <literal>fail_timeout</literal> parameter
after which it will be considered down for a period of time also set
by the <literal>fail_timeout</literal> parameter.
By default, the number of unsuccessful attempts is set to 1.
A value of zero disables accounting of attempts.
What is considered to be an unsuccessful attempt is configured by the
<link doc="ngx_http_proxy_module.xml" id="proxy_next_upstream"/>
and <link doc="ngx_http_fastcgi_module.xml" id="fastcgi_next_upstream"/>
directives.
The <literal>http_404</literal> state is not considered
an unsuccessful attempt.
</tag-desc>

<tag-name><literal>fail_timeout</literal>=<value>time</value></tag-name>
<tag-desc>
sets
<list type="bullet">

<listitem>
a time during which the specified number of unsuccessful attempts to
communicate with the server should happen for the server to be
considered down;
</listitem>

<listitem>
and a period of time the server will be considered down.
</listitem>

</list>
By default, timeout is set to 10 seconds.
</tag-desc>

<tag-name><literal>backup</literal></tag-name>
<tag-desc>
marks the server as a backup server.
It will be passed requests when the primary servers are down.
</tag-desc>

<tag-name><literal>down</literal></tag-name>
<tag-desc>
marks the server as permanently down; used along with
the <link id="ip_hash"/> directive.
</tag-desc>

</list>
</para>

<para>
Example:
<example>
upstream backend {
    server backend1.example.com     weight=5;
    server 127.0.0.1:8080           max_fails=3 fail_timeout=30s;
    server unix:/tmp/backend3;

    server backup1.example.com:8080 backup;
}
</example>
</para>

</directive>


<directive name="upstream">
<syntax block="yes"><value>name</value></syntax>
<default/>
<context>http</context>

<para>
Defines a group of servers.
Servers can listen on different ports.
In addition, servers listening on TCP and UNIX-domain sockets
can be used simultaneously.
</para>

<para>
Example:
<example>
upstream backend {
    server backend1.example.com weight=5;
    server 127.0.0.1:8080       max_fails=3 fail_timeout=30s;
    server unix:/tmp/backend3;
}
</example>
</para>

<para>
Requests are distributed between servers in a round-robin fashion,
taking into account the weights of servers.
In the above example, each 7 requests will be distributed as follows:
5 requests to <literal>backend1.example.com</literal>
and one request to each of second and third servers.
If an error occurs when communicating with the server, a request will
be passed to the next server, and so on until all of the functioning
servers will be tried.
If a successful response could not be obtained from any of the servers,
the client will be returned the result of contacting the last server.
</para>

</directive>

</section>


<section id="variables" name="Embedded Variables">

<para>
The <literal>ngx_http_upstream_module</literal> module
supports the following embedded variables:
<list type="tag">

<tag-name><var>$upstream_addr</var></tag-name>
<tag-desc>
keeps an IP address and port of the server,
or a path to the UNIX-domain socket.
If several servers were contacted during request processing,
their addresses are separated by commas, e.g.
“<literal>192.168.1.1:80, 192.168.1.2:80, unix:/tmp/sock</literal>”.
If an internal redirect from one server group to another happened
using
<header>X-Accel-Redirect</header> or
<link doc="ngx_http_core_module.xml" id="error_page"/>
then these server groups are separated by colons, e.g.
“<literal>192.168.1.1:80, 192.168.1.2:80, unix:/tmp/sock : 192.168.10.1:80, 192.168.10.2:80</literal>”.
</tag-desc>

<tag-name><var>$upstream_response_time</var></tag-name>
<tag-desc>
keeps servers response times in seconds with a milliseconds resolution.
Several responses are also separated by commas and colons.
</tag-desc>

<tag-name><var>$upstream_status</var></tag-name>
<tag-desc>
keeps servers response codes.
Several responses are also separated by commas and colons.
</tag-desc>

<tag-name><var>$upstream_http_...</var></tag-name>
<tag-desc>
keep server response header fields.
For example, the <header>Server</header> response header field
is made available through the <var>$upstream_http_server</var> variable.
Note that only the last server’s response header fields are saved.
</tag-desc>

</list>
</para>

</section>

</module>