Gunicorn "Green Unicorn" is a widely used high-performance Python WSGI UNIX HTTP server, ported from Ruby's Unicorn (Unicorn) project, using the pre-fork worker mode, has a very simple and lightweight to use Features such as resource consumption and high performance.
Install gunicorn
$ sudo apt-get update
$ sudo apt-get install gunicorn
Run gunicorn:
$ gunicorn [OPTIONS] Module name: variable name
The module name is the python file name, which can be the full path + python file name; the variable name is the callable WSGI (Web Server Gateway) in the python file.
Examples:
# filename：test.py def app(environ, start_response): """Simplest possible application object""" data = 'Hello, World!\n' status = '200 OK' response_headers = [ ('Content-type','text/plain'), ('Content-Length', str(len(data))) ] start_response(status, response_headers) return iter([data])
Run the app:
$ gunicorn --workers=2 test:app
Common configuration parameters:
-c CONFIG, --config=CONFIG
Specify a configuration file (py file).
-b BIND, --bind=BIND
Bind with the specified socket.
-D, --daemon
Running the Gunicorn process in the form of a daemon actually puts this service in the background to run.
-w WORKERS, --workers=WORKERS
The number of work processes. As mentioned above, gunicorn is a pre-fork worker mode, which means that when gunicorn starts, a specified number of worker processes will be pre-forked in the main process. When processing requests, gunicorn relies on the operating system to provide load balancing. Workers are usually recommended. The quantity is: (2 x $ num_cores) + 1
-k WORKERCLASS, --worker-class=WORKERCLASS
Worker process type. Includes sync (default), eventlet, gevent, or tornado, gthread, gaiohttp.
--backlog INT
The maximum number of pending connections.
--chdir
Switch to the specified working directory.
--log-level LEVEL
The granularity of the output error log, valid LEVEL are:
debug
info
warning
error
critical
--access-logfile FILE
Confirm the file FILE to be written to the Access log. '-' Means output to standard output.
--error-logfile FILE, --log-file FILE
Confirm the file FILE to be written to the error log. '-' Means output to standard error output.
gunicorn configuration
Gunicorn gets configuration from three different places:
Frame settings (usually only affects Paster applications)
Configuration file (python file): The configuration in the configuration file will override the settings of the framework.
Command Line
The framework settings are only related to Paster (a web framework), not discussed; the command line configuration is shown in the above section; now we see how to configure gunicorn with configuration files:
The configuration file must be a python file, just write the parameters on the command line into the py file. If you need to set which parameter, you can assign a value to the parameter in the py file. E.g:
# example.py bind = "127.0.0.1:8000" workers = 2
Run gunicorn:
$ gunicorn -c example.py test:app
Equivalent to:
$ gunicorn -w 2 -b 127.0.0.1:8000 test:app
Of course, the configuration file can also achieve more complex configuration:
# Gunicorn.py Import the logging Import logging.handlers from logging.handlers Import WatchedFileHandler Import OS Import multiprocessing the bind = ' 127.0.0.1:8000 ' # binding ip and port backlog = 512 # listen queue the chdir = ' / Home / Test / server / bin ' # gunicorn to switch to the target working directory timeout = 30 # timeout worker_class = ' gevent ' # use gevent mode, you can also use sync mode, the default is sync mode = multiprocessing.cpu_count Workers () * 2 + 1 # number of processes Threads = 2 # specifies the number of threads per process open the LogLevel = ' info ' # log level, the log level refers to the level of the error log, and access log The level cannot be set access_log_format = ' % (t) s% (p) s% (h) s "% (r) s"% (s) s% (L) s% (b) s% (f) s "" % (a) S ' ' # set gunicorn access log format, the error log can not be set "" " meaning that each option is as follows: H Remote address L '-' U Currently '-', On May BE User name Future Releases in T date of the request r status line (eg `` GET / HTTP / 1.1``) s status b response length or '-' f referer User Agent A Request Time in seconds The T D Request Time in microseconds L decimal Request Time in seconds The P Process ID "" " the accesslog = " /home/test/server/log/gunicorn_access.log " # access log file errorlog = " / Home / Test / Server / log / gunicorn_error.log " # error log file
Run gunicorn:
$ gunicorn -c gunicorn.py test:app