NAME

   Starman - High-performance preforking PSGI/Plack web server

SYNOPSIS

     # Run app.psgi with the default settings
     > starman

     # run with Server::Starter
     > start_server --port 127.0.0.1:80 -- starman --workers 32 myapp.psgi

     # UNIX domain sockets
     > starman --listen /tmp/starman.sock

   Read more options and configurations by running `perldoc starman`
   (lower-case s).

DESCRIPTION

   Starman is a PSGI perl web server that has unique features such as:

   High Performance

     Uses the fast XS/C HTTP header parser

   Preforking

     Spawns workers preforked like most high performance UNIX servers do.
     Starman also reaps dead children and automatically restarts the
     worker pool.

   Signals

     Supports HUP for graceful worker restarts, and TTIN/TTOU to
     dynamically increase or decrease the number of worker processes, as
     well as QUIT to gracefully shutdown the worker processes.

   Superdaemon aware

     Supports Server::Starter for hot deploy and graceful restarts.

   Multiple interfaces and UNIX Domain Socket support

     Able to listen on multiple interfaces including UNIX sockets.

   Small memory footprint

     Preloading the applications with --preload-app command line option
     enables copy-on-write friendly memory management. Also, the minimum
     memory usage Starman requires for the master process is 7MB and
     children (workers) is less than 3.0MB.

   PSGI compatible

     Can run any PSGI applications and frameworks

   HTTP/1.1 support

     Supports chunked requests and responses, keep-alive and pipeline
     requests.

   UNIX only

     This server does not support Win32.

PERFORMANCE

   Here's a simple benchmark using Hello.psgi.

     -- server: Starman (workers=10)
     Requests per second:    6849.16 [#/sec] (mean)
     -- server: Twiggy
     Requests per second:    3911.78 [#/sec] (mean)
     -- server: AnyEvent::HTTPD
     Requests per second:    2738.49 [#/sec] (mean)
     -- server: HTTP::Server::PSGI
     Requests per second:    2218.16 [#/sec] (mean)
     -- server: HTTP::Server::PSGI (workers=10)
     Requests per second:    2792.99 [#/sec] (mean)
     -- server: HTTP::Server::Simple
     Requests per second:    1435.50 [#/sec] (mean)
     -- server: Corona
     Requests per second:    2332.00 [#/sec] (mean)
     -- server: POE
     Requests per second:    503.59 [#/sec] (mean)

   This benchmark was processed with ab -c 10 -t 1 -k on MacBook Pro 13"
   late 2009 model on Mac OS X 10.6.2 with perl 5.10.0. YMMV.

NOTES

   Because Starman runs as a preforking model, it is not recommended to
   serve the requests directly from the internet, especially when slow
   requesting clients are taken into consideration. It is suggested to put
   Starman workers behind the frontend servers such as nginx, and use HTTP
   proxy with TCP or UNIX sockets.

AUTHOR

   Tatsuhiko Miyagawa <[email protected]>

   Andy Grundman wrote Catalyst::Engine::HTTP::Prefork, which this module
   is heavily based on.

   Kazuho Oku wrote Net::Server::SS::PreFork that makes it easy to add
   Server::Starter support to this software.

COPYRIGHT

   Tatsuhiko Miyagawa, 2010-

LICENSE

   This library is free software; you can redistribute it and/or modify it
   under the same terms as Perl itself.

SEE ALSO

   Plack Catalyst::Engine::HTTP::Prefork Net::Server::PreFork

You are viewing proxied material from ftp.icm.edu.pl. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.