Module ngx_http_core_module

absolute_redirect on;

If disabled, redirects issued by nginx will be relative.

See also server_name_in_redirect and port_in_redirect directives.

aio off;

Enables or disables the use of asynchronous file I/O (AIO) on FreeBSD and Linux:

location /video/ {
    aio            on;
    output_buffers 1 64k;
}

On FreeBSD, AIO can be used starting from FreeBSD 4.3. Prior to FreeBSD 11.0, AIO can either be linked statically into a kernel:

options VFS_AIO

or loaded dynamically as a kernel loadable module:

kldload aio

On Linux, AIO can be used starting from kernel version 2.6.22. Also, it is necessary to enable directio, or otherwise reading will be blocking:

location /video/ {
    aio            on;
    directio       512;
    output_buffers 1 128k;
}

On Linux, directio can only be used for reading blocks that are aligned on 512-byte boundaries (or 4K for XFS). File’s unaligned end is read in blocking mode. The same holds true for byte range requests and for FLV requests not from the beginning of a file: reading of unaligned data at the beginning and end of a file will be blocking.

When both AIO and sendfile are enabled on Linux, AIO is used for files that are larger than or equal to the size specified in the directio directive, while sendfile is used for files of smaller sizes or when directio is disabled.

location /video/ {
    sendfile       on;
    aio            on;
    directio       8m;
}

Finally, files can be read and sent using multi-threading (1.7.11), without blocking a worker process:

location /video/ {
    sendfile       on;
    aio            threads;
}

Read and send file operations are offloaded to threads of the specified pool. If the pool name is omitted, the pool with the name “default” is used. The pool name can also be set with variables:

aio threads=pool$disk;

By default, multi-threading is disabled, it should be enabled with the --with-threads configuration parameter. Currently, multi-threading is compatible only with the epoll, kqueue, and eventport methods. Multi-threaded sending of files is only supported on Linux.

See also the sendfile directive.

aio_write off;

If aio is enabled, specifies whether it is used for writing files. Currently, this only works when using aio threads and is limited to writing temporary files with data received from proxied servers.

Defines a replacement for the specified location. For example, with the following configuration

location /i/ {
    alias /data/w3/images/;
}

on request of “/i/top.gif”, the file /data/w3/images/top.gif will be sent.

The path value can contain variables, except $document_root and $realpath_root.

If alias is used inside a location defined with a regular expression then such regular expression should contain captures and alias should refer to these captures (0.7.40), for example:

location ~ ^/users/(.+\.(?:gif|jpe?g|png))$ {
    alias /data/w3/images/$1;
}

When location matches the last part of the directive’s value:

location /images/ {
    alias /data/w3/images/;
}

it is better to use the root directive instead:

location /images/ {
    root /data/w3;
}

auth_delay 0s;

Delays processing of unauthorized requests with 401 response code to prevent timing attacks when access is limited by password or by the result of subrequest.

chunked_transfer_encoding on;

Allows disabling chunked transfer encoding in HTTP/1.1. It may come in handy when using a software failing to support chunked encoding despite the standard’s requirement.

client_body_buffer_size 8k|16k;

Sets buffer size for reading client request body. In case the request body is larger than the buffer, the whole body or only its part is written to a temporary file. By default, buffer size is equal to two memory pages. This is 8K on x86, other 32-bit platforms, and x86-64. It is usually 16K on other 64-bit platforms.

client_body_in_file_only off;

Determines whether nginx should save the entire client request body into a file. This directive can be used during debugging, or when using the $request_body_file variable, or the $r->request_body_file method of the module ngx_http_perl_module.

When set to the value on, temporary files are not removed after request processing.

The value clean will cause the temporary files left after request processing to be removed.

client_body_in_single_buffer off;

Determines whether nginx should save the entire client request body in a single buffer. The directive is recommended when using the $request_body variable, to save the number of copy operations involved.

client_body_temp_path client_body_temp;

Defines a directory for storing temporary files holding client request bodies. Up to three-level subdirectory hierarchy can be used under the specified directory. For example, in the following configuration

client_body_temp_path /spool/nginx/client_temp 1 2;

a path to a temporary file might look like this:

/spool/nginx/client_temp/7/45/00000123457

client_body_timeout 60s;

Defines a timeout for reading client request body. The timeout is set only for a period between two successive read operations, not for the transmission of the whole request body. If a client does not transmit anything within this time, the request is terminated with the 408 (Request Time-out) error.

client_header_buffer_size 1k;

Sets buffer size for reading client request header. For most requests, a buffer of 1K bytes is enough. However, if a request includes long cookies, or comes from a WAP client, it may not fit into 1K. If a request line or a request header field does not fit into this buffer then larger buffers, configured by the large_client_header_buffers directive, are allocated.

If the directive is specified on the server level, the value from the default server can be used. Details are provided in the “Virtual server selection” section.

client_header_timeout 60s;

Defines a timeout for reading client request header. If a client does not transmit the entire header within this time, the request is terminated with the 408 (Request Time-out) error.

client_max_body_size 1m;

Sets the maximum allowed size of the client request body. If the size in a request exceeds the configured value, the 413 (Request Entity Too Large) error is returned to the client. Please be aware that browsers cannot correctly display this error. Setting size to 0 disables checking of client request body size.

connection_pool_size 256|512;

Allows accurate tuning of per-connection memory allocations. This directive has minimal impact on performance and should not generally be used. By default, the size is equal to 256 bytes on 32-bit platforms and 512 bytes on 64-bit platforms.

Prior to version 1.9.8, the default value was 256 on all platforms.

default_type text/plain;

Defines the default MIME type of a response. Mapping of file name extensions to MIME types can be set with the types directive.

directio off;

Enables the use of the O_DIRECT flag (FreeBSD, Linux), the F_NOCACHE flag (macOS), or the directio() function (Solaris), when reading files that are larger than or equal to the specified size. The directive automatically disables (0.7.15) the use of sendfile for a given request. It can be useful for serving large files:

directio 4m;

or when using aio on Linux.

directio_alignment 512;

Sets the alignment for directio. In most cases, a 512-byte alignment is enough. However, when using XFS under Linux, it needs to be increased to 4K.

disable_symlinks off;

Determines how symbolic links should be treated when opening files:

off

Symbolic links in the pathname are allowed and not checked. This is the default behavior.

on

If any component of the pathname is a symbolic link, access to a file is denied.

if_not_owner

Access to a file is denied if any component of the pathname is a symbolic link, and the link and object that the link points to have different owners.

from=part

When checking symbolic links (parameters on and if_not_owner), all components of the pathname are normally checked. Checking of symbolic links in the initial part of the pathname may be avoided by specifying additionally the from=part parameter. In this case, symbolic links are checked only from the pathname component that follows the specified initial part. If the value is not an initial part of the pathname checked, the whole pathname is checked as if this parameter was not specified at all. If the value matches the whole file name, symbolic links are not checked. The parameter value can contain variables.

Example:

disable_symlinks on from=$document_root;

This directive is only available on systems that have the openat() and fstatat() interfaces. Such systems include modern versions of FreeBSD, Linux, and Solaris.

Parameters on and if_not_owner add a processing overhead.

On systems that do not support opening of directories only for search, to use these parameters it is required that worker processes have read permissions for all directories being checked.

The ngx_http_autoindex_module, ngx_http_random_index_module, and ngx_http_dav_module modules currently ignore this directive.

Defines the URI that will be shown for the specified errors. A uri value can contain variables.

Example:

error_page 404             /404.html;
error_page 500 502 503 504 /50x.html;

This causes an internal redirect to the specified uri with the client request method changed to “GET” (for all methods other than “GET” and “HEAD”).

Furthermore, it is possible to change the response code to another using the “=response” syntax, for example:

error_page 404 =200 /empty.gif;

If an error response is processed by a proxied server or a FastCGI/uwsgi/SCGI/gRPC server, and the server may return different response codes (e.g., 200, 302, 401 or 404), it is possible to respond with the code it returns:

error_page 404 = /404.php;

If there is no need to change URI and method during internal redirection it is possible to pass error processing into a named location:

location / {
    error_page 404 = @fallback;
}

location @fallback {
    proxy_pass http://backend;
}

If uri processing leads to an error, the status code of the last occurred error is returned to the client.

It is also possible to use URL redirects for error processing:

error_page 403      http://example.com/forbidden.html;
error_page 404 =301 http://example.com/notfound.html;

In this case, by default, the response code 302 is returned to the client. It can only be changed to one of the redirect status codes (301, 302, 303, 307, and 308).

The code 307 was not treated as a redirect until versions 1.1.16 and 1.0.13.

The code 308 was not treated as a redirect until version 1.13.0.

These directives are inherited from the previous configuration level if and only if there are no error_page directives defined on the current level.

etag on;

Enables or disables automatic generation of the “ETag” response header field for static resources.

Provides the configuration file context in which the HTTP server directives are specified.

if_modified_since exact;

Specifies how to compare modification time of a response with the time in the “If-Modified-Since” request header field:

off

the response is always considered modified (0.7.34);

exact

exact match;

before

modification time of the response is less than or equal to the time in the “If-Modified-Since” request header field.

ignore_invalid_headers on;

Controls whether header fields with invalid names should be ignored. Valid names are composed of English letters, digits, hyphens, and possibly underscores (as controlled by the underscores_in_headers directive).

If the directive is specified on the server level, the value from the default server can be used. Details are provided in the “Virtual server selection” section.

Specifies that a given location can only be used for internal requests. For external requests, the client error 404 (Not Found) is returned. Internal requests are the following:

• requests redirected by the error_page, index, random_index, and try_files directives;
• requests redirected by the “X-Accel-Redirect” response header field from an upstream server;
• subrequests formed by the “include virtual” command of the ngx_http_ssi_module module, by the ngx_http_addition_module module directives, and by auth_request and mirror directives;
• requests changed by the rewrite directive.

Example:

error_page 404 /404.html;

location = /404.html {
    internal;
}

There is a limit of 10 internal redirects per request to prevent request processing cycles that can occur in incorrect configurations. If this limit is reached, the error 500 (Internal Server Error) is returned. In such cases, the “rewrite or internal redirection cycle” message can be seen in the error log.

keepalive_disable msie6;

Disables keep-alive connections with misbehaving browsers. The browser parameters specify which browsers will be affected. The value msie6 disables keep-alive connections with old versions of MSIE, once a POST request is received. The value safari disables keep-alive connections with Safari and Safari-like browsers on macOS and macOS-like operating systems. The value none enables keep-alive connections with all browsers.

Prior to version 1.1.18, the value safari matched all Safari and Safari-like browsers on all operating systems, and keep-alive connections with them were disabled by default.

keepalive_requests 1000;

Sets the maximum number of requests that can be served through one keep-alive connection. After the maximum number of requests are made, the connection is closed.

Closing connections periodically is necessary to free per-connection memory allocations. Therefore, using too high maximum number of requests could result in excessive memory usage and not recommended.

Prior to version 1.19.10, the default value was 100.

keepalive_time 1h;

Limits the maximum time during which requests can be processed through one keep-alive connection. After this time is reached, the connection is closed following the subsequent request processing.

keepalive_timeout 75s;

The first parameter sets a timeout during which a keep-alive client connection will stay open on the server side. The zero value disables keep-alive client connections. The optional second parameter sets a value in the “Keep-Alive: timeout=time” response header field. Two parameters may differ.

The “Keep-Alive: timeout=time” header field is recognized by Mozilla and Konqueror. MSIE closes keep-alive connections by itself in about 60 seconds.

large_client_header_buffers 4 8k;

Sets the maximum number and size of buffers used for reading large client request header. A request line cannot exceed the size of one buffer, or the 414 (Request-URI Too Large) error is returned to the client. A request header field cannot exceed the size of one buffer as well, or the 400 (Bad Request) error is returned to the client. Buffers are allocated only on demand. By default, the buffer size is equal to 8K bytes. If after the end of request processing a connection is transitioned into the keep-alive state, these buffers are released.

If the directive is specified on the server level, the value from the default server can be used. Details are provided in the “Virtual server selection” section.

Limits allowed HTTP methods inside a location. The method parameter can be one of the following: GET, HEAD, POST, PUT, DELETE, MKCOL, COPY, MOVE, OPTIONS, PROPFIND, PROPPATCH, LOCK, UNLOCK, or PATCH. Allowing the GET method makes the HEAD method also allowed. Access to other methods can be limited using the ngx_http_access_module and ngx_http_auth_basic_module modules directives:

limit_except GET {
    allow 192.168.1.0/32;
    deny  all;
}

Please note that this will limit access to all methods except GET and HEAD.

limit_rate 0;

Limits the rate of response transmission to a client. The rate is specified in bytes per second. The zero value disables rate limiting. The limit is set per a request, and so if a client simultaneously opens two connections, the overall rate will be twice as much as the specified limit.

Parameter value can contain variables (1.17.0). It may be useful in cases where rate should be limited depending on a certain condition:

map $slow $rate {
    1     4k;
    2     8k;
}

limit_rate $rate;

Rate limit can also be set in the $limit_rate variable, however, since version 1.17.0, this method is not recommended:

server {

    if ($slow) {
        set $limit_rate 4k;
    }

    ...
}

Rate limit can also be set in the “X-Accel-Limit-Rate” header field of a proxied server response. This capability can be disabled using the proxy_ignore_headers, fastcgi_ignore_headers, uwsgi_ignore_headers, and scgi_ignore_headers directives.

limit_rate_after 0;

Sets the initial amount after which the further transmission of a response to a client will be rate limited. Parameter value can contain variables (1.17.0).

Example:

location /flv/ {
    flv;
    limit_rate_after 500k;
    limit_rate       50k;
}

lingering_close on;

Controls how nginx closes client connections.

The default value “on” instructs nginx to wait for and process additional data from a client before fully closing a connection, but only if heuristics suggests that a client may be sending more data.

The value “always” will cause nginx to unconditionally wait for and process additional client data.

The value “off” tells nginx to never wait for more data and close the connection immediately. This behavior breaks the protocol and should not be used under normal circumstances.

To control closing HTTP/2 connections, the directive must be specified on the server level (1.19.1).

lingering_time 30s;

When lingering_close is in effect, this directive specifies the maximum time during which nginx will process (read and ignore) additional data coming from a client. After that, the connection will be closed, even if there will be more data.

lingering_timeout 5s;

When lingering_close is in effect, this directive specifies the maximum waiting time for more client data to arrive. If data are not received during this time, the connection is closed. Otherwise, the data are read and ignored, and nginx starts waiting for more data again. The “wait-read-ignore” cycle is repeated, but no longer than specified by the lingering_time directive.

listen *:80 | *:8000;

Sets the address and port for IP, or the path for a UNIX-domain socket on which the server will accept requests. Both address and port, or only address or only port can be specified. An address may also be a hostname, for example:

listen 127.0.0.1:8000;
listen 127.0.0.1;
listen 8000;
listen *:8000;
listen localhost:8000;

IPv6 addresses (0.7.36) are specified in square brackets:

listen [::]:8000;
listen [::1];

UNIX-domain sockets (0.8.21) are specified with the “unix:” prefix:

listen unix:/var/run/nginx.sock;

If only address is given, the port 80 is used.

If the directive is not present then either *:80 is used if nginx runs with the superuser privileges, or *:8000 otherwise.

The default_server parameter, if present, will cause the server to become the default server for the specified address:port pair. If none of the directives have the default_server parameter then the first server with the address:port pair will be the default server for this pair.

In versions prior to 0.8.21 this parameter is named simply default.

The ssl parameter (0.7.14) allows specifying that all connections accepted on this port should work in SSL mode. This allows for a more compact configuration for the server that handles both HTTP and HTTPS requests.

The http2 parameter (1.9.5) configures the port to accept HTTP/2 connections. Normally, for this to work the ssl parameter should be specified as well, but nginx can also be configured to accept HTTP/2 connections without SSL.

The parameter is deprecated, the http2 directive should be used instead.

The quic parameter (1.25.0) configures the port to accept QUIC connections.

The proxy_protocol parameter (1.5.12) allows specifying that all connections accepted on this port should use the PROXY protocol.

The PROXY protocol version 2 is supported since version 1.13.11.

The listen directive can have several additional parameters specific to socket-related system calls. These parameters can be specified in any listen directive, but only once for a given address:port pair.

In versions prior to 0.8.21, they could only be specified in the listen directive together with the default parameter.

setfib=number

this parameter (0.8.44) sets the associated routing table, FIB (the SO_SETFIB option) for the listening socket. This currently works only on FreeBSD.

fastopen=number

enables “TCP Fast Open” for the listening socket (1.5.8) and limits the maximum length for the queue of connections that have not yet completed the three-way handshake.

Do not enable this feature unless the server can handle receiving the same SYN packet with data more than once.

backlog=number

sets the backlog parameter in the listen() call that limits the maximum length for the queue of pending connections. By default, backlog is set to -1 on FreeBSD, DragonFly BSD, and macOS, and to 511 on other platforms.

rcvbuf=size

sets the receive buffer size (the SO_RCVBUF option) for the listening socket.

sndbuf=size

sets the send buffer size (the SO_SNDBUF option) for the listening socket.

accept_filter=filter

sets the name of accept filter (the SO_ACCEPTFILTER option) for the listening socket that filters incoming connections before passing them to accept(). This works only on FreeBSD and NetBSD 5.0+. Possible values are dataready and httpready.

deferred

instructs to use a deferred accept() (the TCP_DEFER_ACCEPT socket option) on Linux.

bind

instructs to make a separate bind() call for a given address:port pair. This is useful because if there are several listen directives with the same port but different addresses, and one of the listen directives listens on all addresses for the given port (*:port), nginx will bind() only to *:port. It should be noted that the getsockname() system call will be made in this case to determine the address that accepted the connection. If the setfib, fastopen, backlog, rcvbuf, sndbuf, accept_filter, deferred, ipv6only, reuseport, or so_keepalive parameters are used then for a given address:port pair a separate bind() call will always be made.

ipv6only=on|off

this parameter (0.7.42) determines (via the IPV6_V6ONLY socket option) whether an IPv6 socket listening on a wildcard address [::] will accept only IPv6 connections or both IPv6 and IPv4 connections. This parameter is turned on by default. It can only be set once on start.

Prior to version 1.3.4, if this parameter was omitted then the operating system’s settings were in effect for the socket.

reuseport

this parameter (1.9.1) instructs to create an individual listening socket for each worker process (using the SO_REUSEPORT socket option on Linux 3.9+ and DragonFly BSD, or SO_REUSEPORT_LB on FreeBSD 12+), allowing a kernel to distribute incoming connections between worker processes. This currently works only on Linux 3.9+, DragonFly BSD, and FreeBSD 12+ (1.15.1).

Inappropriate use of this option may have its security implications.

so_keepalive=on|off|[keepidle]:[keepintvl]:[keepcnt]

this parameter (1.1.11) configures the “TCP keepalive” behavior for the listening socket. If this parameter is omitted then the operating system’s settings will be in effect for the socket. If it is set to the value “on”, the SO_KEEPALIVE option is turned on for the socket. If it is set to the value “off”, the SO_KEEPALIVE option is turned off for the socket. Some operating systems support setting of TCP keepalive parameters on a per-socket basis using the TCP_KEEPIDLE, TCP_KEEPINTVL, and TCP_KEEPCNT socket options. On such systems (currently, Linux 2.4+, NetBSD 5+, and FreeBSD 9.0-STABLE), they can be configured using the keepidle, keepintvl, and keepcnt parameters. One or two parameters may be omitted, in which case the system default setting for the corresponding socket option will be in effect. For example,

so_keepalive=30m::10

will set the idle timeout (TCP_KEEPIDLE) to 30 minutes, leave the probe interval (TCP_KEEPINTVL) at its system default, and set the probes count (TCP_KEEPCNT) to 10 probes.

Example:

listen 127.0.0.1 default_server accept_filter=dataready backlog=1024;

Sets configuration depending on a request URI.

The matching is performed against a normalized URI, after decoding the text encoded in the “%XX” form, resolving references to relative path components “.” and “..”, and possible compression of two or more adjacent slashes into a single slash.

A location can either be defined by a prefix string, or by a regular expression. Regular expressions are specified with the preceding “~*” modifier (for case-insensitive matching), or the “~” modifier (for case-sensitive matching). To find location matching a given request, nginx first checks locations defined using the prefix strings (prefix locations). Among them, the location with the longest matching prefix is selected and remembered. Then regular expressions are checked, in the order of their appearance in the configuration file. The search of regular expressions terminates on the first match, and the corresponding configuration is used. If no match with a regular expression is found then the configuration of the prefix location remembered earlier is used.

location blocks can be nested, with some exceptions mentioned below.

For case-insensitive operating systems such as macOS and Cygwin, matching with prefix strings ignores a case (0.7.7). However, comparison is limited to one-byte locales.

Regular expressions can contain captures (0.7.40) that can later be used in other directives.

If the longest matching prefix location has the “^~” modifier then regular expressions are not checked.

Also, using the “=” modifier it is possible to define an exact match of URI and location. If an exact match is found, the search terminates. For example, if a “/” request happens frequently, defining “location = /” will speed up the processing of these requests, as search terminates right after the first comparison. Such a location cannot obviously contain nested locations.

In versions from 0.7.1 to 0.8.41, if a request matched the prefix location without the “=” and “^~” modifiers, the search also terminated and regular expressions were not checked.

Let’s illustrate the above by an example:

location = / {
    [ configuration A ]
}

location / {
    [ configuration B ]
}

location /documents/ {
    [ configuration C ]
}

location ^~ /images/ {
    [ configuration D ]
}

location ~* \.(gif|jpg|jpeg)$ {
    [ configuration E ]
}

The “/” request will match configuration A, the “/index.html” request will match configuration B, the “/documents/document.html” request will match configuration C, the “/images/1.gif” request will match configuration D, and the “/documents/1.jpg” request will match configuration E.

The “@” prefix defines a named location. Such a location is not used for a regular request processing, but instead used for request redirection. They cannot be nested, and cannot contain nested locations.

If a location is defined by a prefix string that ends with the slash character, and requests are processed by one of proxy_pass, fastcgi_pass, uwsgi_pass, scgi_pass, memcached_pass, or grpc_pass, then the special processing is performed. In response to a request with URI equal to this string, but without the trailing slash, a permanent redirect with the code 301 will be returned to the requested URI with the slash appended. If this is not desired, an exact match of the URI and location could be defined like this:

location /user/ {
    proxy_pass http://user.example.com;
}

location = /user {
    proxy_pass http://login.example.com;
}

log_not_found on;

Enables or disables logging of errors about not found files into error_log.

log_subrequest off;

Enables or disables logging of subrequests into access_log.

max_headers 1000;

Limits the maximum allowed number of client request header fields. If the limit is exceeded, the 400 (Bad Request) error is returned to the client.

If the directive is specified on the server level, the value from the default server can be used. Details are provided in the “Virtual server selection” section.

Limits the maximum allowed number of ranges in byte-range requests. Requests that exceed the limit are processed as if there were no byte ranges specified. By default, the number of ranges is not limited. The zero value disables the byte-range support completely.

merge_slashes on;

Enables or disables compression of two or more adjacent slashes in a URI into a single slash.

Note that compression is essential for the correct matching of prefix string and regular expression locations. Without it, the “//scripts/one.php” request would not match

location /scripts/ {
    ...
}

and might be processed as a static file. So it gets converted to “/scripts/one.php”.

Turning the compression off can become necessary if a URI contains base64-encoded names, since base64 uses the “/” character internally. However, for security considerations, it is better to avoid turning the compression off.

If the directive is specified on the server level, the value from the default server can be used. Details are provided in the “Virtual server selection” section.

msie_padding on;

Enables or disables adding comments to responses for MSIE clients with status greater than 400 to increase the response size to 512 bytes.

msie_refresh off;

Enables or disables issuing refreshes instead of redirects for MSIE clients.

open_file_cache off;

Configures a cache that can store:

• open file descriptors, their sizes and modification times;
• information on existence of directories;
• file lookup errors, such as “file not found”, “no read permission”, and so on.

  Caching of errors should be enabled separately by the open_file_cache_errors directive.

The directive has the following parameters:

max

sets the maximum number of elements in the cache; on cache overflow the least recently used (LRU) elements are removed;

inactive

defines a time after which an element is removed from the cache if it has not been accessed during this time; by default, it is 60 seconds;

off

disables the cache.

Example:

open_file_cache          max=1000 inactive=20s;
open_file_cache_valid    30s;
open_file_cache_min_uses 2;
open_file_cache_errors   on;

open_file_cache_errors off;

Enables or disables caching of file lookup errors by open_file_cache.

open_file_cache_min_uses 1;

Sets the minimum number of file accesses during the period configured by the inactive parameter of the open_file_cache directive, required for a file descriptor to remain open in the cache.

open_file_cache_valid 60s;

Sets a time after which open_file_cache elements should be validated.

output_buffers 2 32k;

Sets the number and size of the buffers used for reading a response from a disk.

Prior to version 1.9.5, the default value was 1 32k.

port_in_redirect on;

Enables or disables specifying the port in absolute redirects issued by nginx.

The use of the primary server name in redirects is controlled by the server_name_in_redirect directive.

postpone_output 1460;

If possible, the transmission of client data will be postponed until nginx has at least size bytes of data to send. The zero value disables postponing data transmission.

read_ahead 0;

Sets the amount of pre-reading for the kernel when working with file.

On Linux, the posix_fadvise(0, 0, 0, POSIX_FADV_SEQUENTIAL) system call is used, and so the size parameter is ignored.

On FreeBSD, the fcntl(O_READAHEAD, size) system call, supported since FreeBSD 9.0-CURRENT, is used. FreeBSD 7 has to be patched.

recursive_error_pages off;

Enables or disables doing several redirects using the error_page directive. The number of such redirects is limited.

request_pool_size 4k;

Allows accurate tuning of per-request memory allocations. This directive has minimal impact on performance and should not generally be used.

reset_timedout_connection off;

Enables or disables resetting timed out connections and connections closed with the non-standard code 444 (1.15.2). The reset is performed as follows. Before closing a socket, the SO_LINGER option is set on it with a timeout value of 0. When the socket is closed, TCP RST is sent to the client, and all memory occupied by this socket is released. This helps avoid keeping an already closed socket with filled buffers in a FIN_WAIT1 state for a long time.

It should be noted that timed out keep-alive connections are closed normally.

Configures name servers used to resolve names of upstream servers into addresses, for example:

resolver 127.0.0.1 [::1]:5353;

The address can be specified as a domain name or IP address, with an optional port (1.3.1, 1.2.2). If port is not specified, the port 53 is used. Name servers are queried in a round-robin fashion.

Before version 1.1.7, only a single name server could be configured. Specifying name servers using IPv6 addresses is supported starting from versions 1.3.1 and 1.2.2.

By default, nginx will look up both IPv4 and IPv6 addresses while resolving. If looking up of IPv4 or IPv6 addresses is not desired, the ipv4=off (1.23.1) or the ipv6=off parameter can be specified.

Resolving of names into IPv6 addresses is supported starting from version 1.5.8.

By default, nginx caches answers using the TTL value of a response. An optional valid parameter allows overriding it:

resolver 127.0.0.1 [::1]:5353 valid=30s;

Before version 1.1.9, tuning of caching time was not possible, and nginx always cached answers for the duration of 5 minutes.

To prevent DNS spoofing, it is recommended configuring DNS servers in a properly secured trusted local network.

resolver_timeout 30s;

Sets a timeout for name resolution, for example:

resolver_timeout 5s;

root html;

Sets the root directory for requests. For example, with the following configuration

location /i/ {
    root /data/w3;
}

The /data/w3/i/top.gif file will be sent in response to the “/i/top.gif” request.

The path value can contain variables, except $document_root and $realpath_root.

A path to the file is constructed by merely adding a URI to the value of the root directive. If a URI has to be modified, the alias directive should be used.

satisfy all;

Allows access if all (all) or at least one (any) of the ngx_http_access_module, ngx_http_auth_basic_module, or ngx_http_auth_request_module modules allow access.

Example:

location / {
    satisfy any;

    allow 192.168.1.0/32;
    deny  all;

    auth_basic           "closed site";
    auth_basic_user_file conf/htpasswd;
}

send_lowat 0;

If the directive is set to a non-zero value, nginx will try to minimize the number of send operations on client sockets by using either NOTE_LOWAT flag of the kqueue method or the SO_SNDLOWAT socket option. In both cases the specified size is used.

This directive is ignored on Linux, Solaris, and Windows.

send_timeout 60s;

Sets a timeout for transmitting a response to the client. The timeout is set only between two successive write operations, not for the transmission of the whole response. If the client does not receive anything within this time, the connection is closed.

sendfile off;

Enables or disables the use of sendfile().

Starting from nginx 0.8.12 and FreeBSD 5.2.1, aio can be used to pre-load data for sendfile():

location /video/ {
    sendfile       on;
    tcp_nopush     on;
    aio            on;
}

In this configuration, sendfile() is called with the SF_NODISKIO flag which causes it not to block on disk I/O, but, instead, report back that the data are not in memory. nginx then initiates an asynchronous data load by reading one byte. On the first read, the FreeBSD kernel loads the first 128K bytes of a file into memory, although next reads will only load data in 16K chunks. This can be changed using the read_ahead directive.

Before version 1.7.11, pre-loading could be enabled with aio sendfile;.

sendfile_max_chunk 2m;

Limits the amount of data that can be transferred in a single sendfile() call. Without the limit, one fast connection may seize the worker process entirely.

Prior to version 1.21.4, by default there was no limit.

Sets configuration for a virtual server. There is no clear separation between IP-based (based on the IP address) and name-based (based on the “Host” request header field) virtual servers. Instead, the listen directives describe all addresses and ports that should accept connections for the server, and the server_name directive lists all server names. Example configurations are provided in the “How nginx processes a request” document.

server_name "";

Sets names of a virtual server, for example:

server {
    server_name example.com www.example.com;
}

The first name becomes the primary server name.

Server names can include an asterisk (“*”) replacing the first or last part of a name:

server {
    server_name example.com *.example.com www.example.*;
}

Such names are called wildcard names.

The first two of the names mentioned above can be combined in one:

server {
    server_name .example.com;
}

It is also possible to use regular expressions in server names, preceding the name with a tilde (“~”):

server {
    server_name www.example.com ~^www\d+\.example\.com$;
}

Regular expressions can contain captures (0.7.40) that can later be used in other directives:

server {
    server_name ~^(www\.)?(.+)$;

    location / {
        root /sites/$2;
    }
}

server {
    server_name _;

    location / {
        root /sites/default;
    }
}

Named captures in regular expressions create variables (0.8.25) that can later be used in other directives:

server {
    server_name ~^(www\.)?(?<domain>.+)$;

    location / {
        root /sites/$domain;
    }
}

server {
    server_name _;

    location / {
        root /sites/default;
    }
}

If the directive’s parameter is set to “$hostname” (0.9.4), the machine’s hostname is inserted.

It is also possible to specify an empty server name (0.7.11):

server {
    server_name www.example.com "";
}

It allows this server to process requests without the “Host” header field — instead of the default server — for the given address:port pair. This is the default setting.

Before 0.8.48, the machine’s hostname was used by default.

During searching for a virtual server by name, if the name matches more than one of the specified variants, (e.g. both a wildcard name and regular expression match), the first matching variant will be chosen, in the following order of priority:

1. the exact name
2. the longest wildcard name starting with an asterisk, e.g. “*.example.com”
3. the longest wildcard name ending with an asterisk, e.g. “mail.*”
4. the first matching regular expression (in order of appearance in the configuration file)

Detailed description of server names is provided in a separate Server names document.

server_name_in_redirect off;

Enables or disables the use of the primary server name, specified by the server_name directive, in absolute redirects issued by nginx. When the use of the primary server name is disabled, the name from the “Host” request header field is used. If this field is not present, the IP address of the server is used.

The use of a port in redirects is controlled by the port_in_redirect directive.

server_names_hash_bucket_size 32|64|128;

Sets the bucket size for the server names hash tables. The default value depends on the size of the processor’s cache line. The details of setting up hash tables are provided in a separate document.

server_names_hash_max_size 512;

Sets the maximum size of the server names hash tables. The details of setting up hash tables are provided in a separate document.

server_tokens on;

Enables or disables emitting nginx version on error pages and in the “Server” response header field.

The build parameter (1.11.10) enables emitting a build name along with nginx version.

subrequest_output_buffer_size 4k|8k;

Sets the size of the buffer used for storing the response body of a subrequest. By default, the buffer size is equal to one memory page. This is either 4K or 8K, depending on a platform. It can be made smaller, however.

The directive is applicable only for subrequests with response bodies saved into memory. For example, such subrequests are created by SSI.

tcp_nodelay on;

Enables or disables the use of the TCP_NODELAY option. The option is enabled when a connection is transitioned into the keep-alive state. Additionally, it is enabled on SSL connections, for unbuffered proxying, and for WebSocket proxying.

tcp_nopush off;

Enables or disables the use of the TCP_NOPUSH socket option on FreeBSD or the TCP_CORK socket option on Linux. The options are enabled only when sendfile is used. Enabling the option allows

• sending the response header and the beginning of a file in one packet, on Linux and FreeBSD 4.*;
• sending a file in full packets.

Checks the existence of files in the specified order and uses the first found file for request processing; the processing is performed in the current context. The path to a file is constructed from the file parameter according to the root and alias directives. It is possible to check directory’s existence by specifying a slash at the end of a name, e.g. “$uri/”. If none of the files were found, an internal redirect to the uri specified in the last parameter is made. For example:

location /images/ {
    try_files $uri /images/default.gif;
}

location = /images/default.gif {
    expires 30s;
}

The last parameter can also point to a named location, as shown in examples below. Starting from version 0.7.51, the last parameter can also be a code:

location / {
    try_files $uri $uri/index.html $uri.html =404;
}

Example in proxying Mongrel:

location / {
    try_files /system/maintenance.html
              $uri $uri/index.html $uri.html
              @mongrel;
}

location @mongrel {
    proxy_pass http://mongrel;
}

Example for Drupal/FastCGI:

location / {
    try_files $uri $uri/ @drupal;
}

location ~ \.php$ {
    try_files $uri @drupal;

    fastcgi_pass ...;

    fastcgi_param SCRIPT_FILENAME /path/to$fastcgi_script_name;
    fastcgi_param SCRIPT_NAME     $fastcgi_script_name;
    fastcgi_param QUERY_STRING    $args;

    ... other fastcgi_param's
}

location @drupal {
    fastcgi_pass ...;

    fastcgi_param SCRIPT_FILENAME /path/to/index.php;
    fastcgi_param SCRIPT_NAME     /index.php;
    fastcgi_param QUERY_STRING    q=$uri&$args;

    ... other fastcgi_param's
}

In the following example,

location / {
    try_files $uri $uri/ @drupal;
}

the try_files directive is equivalent to

location / {
    error_page 404 = @drupal;
    log_not_found off;
}

And here,

location ~ \.php$ {
    try_files $uri @drupal;

    fastcgi_pass ...;

    fastcgi_param SCRIPT_FILENAME /path/to$fastcgi_script_name;

    ...
}

try_files checks the existence of the PHP file before passing the request to the FastCGI server.

Example for Wordpress and Joomla:

location / {
    try_files $uri $uri/ @wordpress;
}

location ~ \.php$ {
    try_files $uri @wordpress;

    fastcgi_pass ...;

    fastcgi_param SCRIPT_FILENAME /path/to$fastcgi_script_name;
    ... other fastcgi_param's
}

location @wordpress {
    fastcgi_pass ...;

    fastcgi_param SCRIPT_FILENAME /path/to/index.php;
    ... other fastcgi_param's
}

types {
    text/html  html;
    image/gif  gif;
    image/jpeg jpg;
}

Maps file name extensions to MIME types of responses. Extensions are case-insensitive. Several extensions can be mapped to one type, for example:

types {
    application/octet-stream bin exe dll;
    application/octet-stream deb;
    application/octet-stream dmg;
}

A sufficiently full mapping table is distributed with nginx in the conf/mime.types file.

To make a particular location emit the “application/octet-stream” MIME type for all requests, the following configuration can be used:

location /download/ {
    types        { }
    default_type application/octet-stream;
}

types_hash_bucket_size 64;

Sets the bucket size for the types hash tables. The details of setting up hash tables are provided in a separate document.

Prior to version 1.5.13, the default value depended on the size of the processor’s cache line.

types_hash_max_size 1024;

Sets the maximum size of the types hash tables. The details of setting up hash tables are provided in a separate document.

underscores_in_headers off;

Enables or disables the use of underscores in client request header fields. When the use of underscores is disabled, request header fields whose names contain underscores are marked as invalid and become subject to the ignore_invalid_headers directive.

If the directive is specified on the server level, the value from the default server can be used. Details are provided in the “Virtual server selection” section.

variables_hash_bucket_size 64;

Sets the bucket size for the variables hash table. The details of setting up hash tables are provided in a separate document.

variables_hash_max_size 1024;

Sets the maximum size of the variables hash table. The details of setting up hash tables are provided in a separate document.

Prior to version 1.5.13, the default value was 512.

The ngx_http_core_module module supports embedded variables with names matching the Apache Server variables. First of all, these are variables representing client request header fields, such as $http_user_agent, $http_cookie, and so on. Also there are other variables:

$arg_name

argument name in the request line

$args

arguments in the request line

$binary_remote_addr

client address in a binary form, value’s length is always 4 bytes for IPv4 addresses or 16 bytes for IPv6 addresses

$body_bytes_sent

number of bytes sent to a client, not counting the response header; this variable is compatible with the “%B” parameter of the mod_log_config Apache module

$bytes_sent

number of bytes sent to a client (1.3.8, 1.2.5)

$connection

connection serial number (1.3.8, 1.2.5)

$connection_requests

current number of requests made through a connection (1.3.8, 1.2.5)

$connection_time

connection time in seconds with a milliseconds resolution (1.19.10)

$content_length

“Content-Length” request header field

$content_type

“Content-Type” request header field

$cookie_name

the name cookie

$document_root

root or alias directive’s value for the current request

$document_uri

same as $uri

$host

in this order of precedence: host name from the request line, or host name from the “Host” request header field, or the server name matching a request

$hostname

host name

$http_name

arbitrary request header field; the last part of a variable name is the field name converted to lower case with dashes replaced by underscores

$https

“on” if connection operates in SSL mode, or an empty string otherwise

$is_args

“?” if a request line has arguments, or an empty string otherwise

$limit_rate

setting this variable enables response rate limiting; see limit_rate

$msec

current time in seconds with the milliseconds resolution (1.3.9, 1.2.6)

$nginx_version

nginx version

$pid

PID of the worker process

$pipe

“p” if request was pipelined, “.” otherwise (1.3.12, 1.2.7)

$proxy_protocol_addr

client address from the PROXY protocol header (1.5.12)

The PROXY protocol must be previously enabled by setting the proxy_protocol parameter in the listen directive.

$proxy_protocol_port

client port from the PROXY protocol header (1.11.0)

The PROXY protocol must be previously enabled by setting the proxy_protocol parameter in the listen directive.

$proxy_protocol_server_addr

server address from the PROXY protocol header (1.17.6)

The PROXY protocol must be previously enabled by setting the proxy_protocol parameter in the listen directive.

$proxy_protocol_server_port

server port from the PROXY protocol header (1.17.6)

The PROXY protocol must be previously enabled by setting the proxy_protocol parameter in the listen directive.

$proxy_protocol_tlv_name

TLV from the PROXY Protocol header (1.23.2). The name can be a TLV type name or its numeric value. In the latter case, the value is hexadecimal and should be prefixed with 0x:

$proxy_protocol_tlv_alpn
$proxy_protocol_tlv_0x01

SSL TLVs can also be accessed by TLV type name or its numeric value, both prefixed by ssl_:

$proxy_protocol_tlv_ssl_version
$proxy_protocol_tlv_ssl_0x21

The following TLV type names are supported:

• alpn (0x01) - upper layer protocol used over the connection
• authority (0x02) - host name value passed by the client
• unique_id (0x05) - unique connection id
• netns (0x30) - name of the namespace
• ssl (0x20) - binary SSL TLV structure

The following SSL TLV type names are supported:

• ssl_version (0x21) - SSL version used in client connection
• ssl_cn (0x22) - SSL certificate Common Name
• ssl_cipher (0x23) - name of the used cipher
• ssl_sig_alg (0x24) - algorithm used to sign the certificate
• ssl_key_alg (0x25) - public-key algorithm

Also, the following special SSL TLV type name is supported:

• ssl_verify - client SSL certificate verification result, 0 if the client presented a certificate and it was successfully verified, non-zero otherwise.

The PROXY protocol must be previously enabled by setting the proxy_protocol parameter in the listen directive.

$query_string

same as $args

$realpath_root

an absolute pathname corresponding to the root or alias directive’s value for the current request, with all symbolic links resolved to real paths

$remote_addr

client address

$remote_port

client port

$remote_user

user name supplied with the Basic authentication

$request

full original request line

$request_body

request body

The variable’s value is made available in locations processed by the proxy_pass, fastcgi_pass, uwsgi_pass, and scgi_pass directives when the request body was read to a memory buffer.

$request_body_file

name of a temporary file with the request body

At the end of processing, the file needs to be removed. To always write the request body to a file, client_body_in_file_only needs to be enabled. When the name of a temporary file is passed in a proxied request or in a request to a FastCGI/uwsgi/SCGI server, passing the request body should be disabled by the proxy_pass_request_body off, fastcgi_pass_request_body off, uwsgi_pass_request_body off, or scgi_pass_request_body off directives, respectively.

$request_completion

“OK” if a request has completed, or an empty string otherwise

$request_filename

file path for the current request, based on the root or alias directives, and the request URI

$request_id

unique request identifier generated from 16 random bytes, in hexadecimal (1.11.0)

$request_length

request length (including request line, header, and request body) (1.3.12, 1.2.7)

$request_method

request method, usually “GET” or “POST”

$request_time

request processing time in seconds with a milliseconds resolution (1.3.9, 1.2.6); time elapsed since the first bytes were read from the client

$request_uri

full original request URI (with arguments)

$scheme

request scheme, “http” or “https”

$sent_http_name

arbitrary response header field; the last part of a variable name is the field name converted to lower case with dashes replaced by underscores

$sent_trailer_name

arbitrary field sent at the end of the response (1.13.2); the last part of a variable name is the field name converted to lower case with dashes replaced by underscores

$server_addr

an address of the server which accepted a request

Computing a value of this variable usually requires one system call. To avoid a system call, the listen directives must specify addresses and use the bind parameter.

$server_name

name of the server which accepted a request

$server_port

port of the server which accepted a request

$server_protocol

request protocol, usually “HTTP/1.0”, “HTTP/1.1”, “HTTP/2.0”, or “HTTP/3.0”

$status

response status (1.3.2, 1.2.2)

$tcpinfo_rtt, $tcpinfo_rttvar, $tcpinfo_snd_cwnd, $tcpinfo_rcv_space

information about the client TCP connection; available on systems that support the TCP_INFO socket option

$time_iso8601

local time in the ISO 8601 standard format (1.3.12, 1.2.7)

$time_local

local time in the Common Log Format (1.3.12, 1.2.7)

$uri

current URI in request, normalized

The value of $uri may change during request processing, e.g. when doing internal redirects, or when using index files.