These are the features that control basic Nginx functionality. This page was previously called NginxHttpMainModule.
Syntax: daemon on | off
Do not use the daemon or master_process directives in a production mode; these options are used for development only. You can use
daemon off safely in production mode with runit/daemontools, however you can't do a graceful upgrade.
master_process off should never be used in production.
Syntax: env VAR|VAR=VALUE
The instruction allows to limit a set of variables of environment, to change it values or to create new variables for following cases:
- inheritance of variables during upgrading the binary with zero downtime ;
- for use by the embedded Perl module
- for use by working processes. However it is necessary to keep in mind, that management of behaviour of system libraries in a similar way probably not always as frequently libraries use variables only during initialization, that is still before they can be set by means of the given instruction. Exception to it is the above described updating an executed file with zero downtime.
If variable TZ is not described obviously it is always inherited and is always accessible to the embedded Perl module.
Example of use:
By default, nginx wipes all its environment variables except TZ variable.
- "env NAME" allows to keep NAME variable value got from parent process, i.e., shell.
- "env NAME=val" sets NAME variable value.
Syntax: debug_points [stop | abort]
There are some assertion points inside nginx that allow to stop nginx to attach the debugger, or to abort and to create the core file.
Syntax: error_log file [ debug | info | notice | warn | error | crit ]
Specifies the file where server (and fastcgi) errors are logged.
Default values for the error level:
- in the main section -
- in the HTTP section -
- in the server section -
Nginx supports separate error logging per virtual host. This is a unique feature, which lighttpd refuses to implement. For an example of separate error logging per server, see SeparateErrorLoggingPerVirtualHost and this mailing list thread on separating error logging per virtual host.
If you've built Nginx with
--with-debug, you may also use:
error_log LOGFILE [debug_core | debug_alloc | debug_mutex | debug_event | debug_http | debug_imap];
error_log off does not disable logging - the log will be written to a file named "off". To disable logging, you may use:
error_log /dev/null crit;
Also note that as of version 0.7.53, nginx will use a compiled-in default error log location until it has read the config file. If the user running nginx doesn't have write permission to this log location, nginx will raise an alert like this:
[alert]: could not open error log file: open() "/var/log/nginx/error.log" failed (13: Permission denied)
Syntax: include file | *
You can include any configuration files for what ever purpose you want.
Since 0.4.4, the
include directive also supports filename globbing:
Note that until version 0.6.7, paths are relative to what was specified to
configure via the
--prefix=<PATH> directive, which by default is
/usr/local/nginx. If you didn't set this when you compiled Nginx, then use absolute paths.
Since version 0.6.7, paths are relative to directory of nginx configuration file nginx.conf, but not to nginx prefix directory.
Syntax: lock_file file
Default: compile-time option
nginx uses accept mutex to serialize accept() syscalls. If nginx is built by gcc, Intel C++, or SunPro C++ compilers on i386, amd64, sparc64, and ppc64, then nginx uses the atomic instructions to implement the mutex. In other cases the lock file would be used.
Syntax: master_process on | off
Do not use the "daemon" and "master_process" directives in a production mode, these options are mainly used for development only.
Syntax: pid file
Default: compile-time option
The pid-file. It can be used for the kill-command to send signals to nginx, eg: to reload the config:
kill -HUP `cat /var/log/nginx.pid`
Syntax: ssl_engine engine
Default: system dependent
Here you can set your preferred openssl engine if any available. You can figure out which one do you have with the commandline tool:
openssl engine -t
$ openssl engine -t (cryptodev) BSD cryptodev engine [ available ] (dynamic) Dynamic engine loading support [ unavailable ]
Syntax: timer_resolution t
The directive allows to decrease number gettimeofday() syscalls. By default gettimeofday() is called after each return from kevent(), epoll, /dev/poll, select(), poll().
But if you need an exact time in logs when logging $upstream_response_time, or $msec variables, then you should use
Syntax: user user [group]
Default: nobody nobody
If the master process is run as root, then nginx will setuid()/setgid() to USER/GROUP. If GROUP is not specified, then nginx uses the same name as USER. By default it's
nobody user and
nogroup group or the
--group=GROUP from the
user www users;
Syntax: worker_cpu_affinity cpumask [cpumask...]
With this option you can bind the worker process to a CPU, it calls sched_setaffinity().
Bind each worker process to one CPU only.
Bind the first worker to CPU0/CPU2, bind the second worker to CPU1/CPU3. This is suitable for HTT.
Syntax: worker_priority [-] number
With this option you can give to all worker processes the priority (nice) you need/wish, it calls setpriority().
Syntax: worker_processes number
A worker process is a single-threaded process.
If Nginx is doing CPU-intensive work such as SSL or gzipping and you have 2 or more CPUs/cores, then you may set worker_processes to be equal to the number of CPUs or cores.
If you are serving a lot of static files and the total size of the files is bigger than the available memory, then you may increase worker_processes to fully utilize disk bandwidth.
Your OS may schedule all workers on single CPU/core this can be avoided using worker_cpu_affinity.
Nginx has the ability to use more than one worker process for several reasons:
- to use SMP
- to decrease latency when workers blockend on disk I/O
- to limit number of connections per process when select()/poll() is used
worker_connections from the event sections allows you to calculate
max_clients = worker_processes * worker_connections
Syntax: worker_rlimit_core size
Maximum size of core file per worker. Used for Debugging Nginx.
Syntax: worker_rlimit_nofile limit
Default: No value specified, so OS default.
Specifies the value for maximum file descriptors that can be opened by this process.
Syntax: worker_rlimit_sigpending limit
(Since Linux 2.6.8) Specifies the limit on the number of signals that may be queued for the real user ID of the calling process.
Syntax: working_directory path
This is the working directory for the workers. It's used for core files only and Debugging Nginx. nginx uses absolute paths only, all relative paths in configuration files are relative to
The process ID;