Fastcgi Container

Fastcgi Container: a framework for development of high performance web applications in C++

lpre
26
8
C++

Fastcgi Container

Fastcgi Container is a framework for development of high performance web applications in C++.

Fastcgi Container is a branch of Yandex’s Fastcgi Daemon.

What’s new compared to Fastcgi Daemon:

  • The framework is written in C++11 and does not depend on Boost libraries anymore
  • Support for request filters
  • Support for servlets as an extensions of request handlers
  • Support for sessions
  • Support for authentication and authorization
  • Support for security contexts
  • The framework provides Page Compiler - a command-line C++ server page compiler which generates C++ servlets from JSP-like source files
  • The framework is using CMake as a build system

All filters (including authentication/authorization filter) are executed under the FastCGI role “RESPONDER” and do not require that the roles “FILTER” and/or “AUTHORIZER” are supported by FastCGI connector (e.g. corresponding modules for Apache HTTPD).

Session manager and security facility (support for authentication and authorization) are implemented as an optional pluggable modules: you may load them to your application and use it, or run the application without these features.

Note that form-based authentication depends on session management: if you decide to use form-based authentication, you have to activate the session management as well.

License

LGPL v.3.0, except Page Compiler which is covered by GPL v.3.0.

Requirements

  • A C11 compliant compiler with complete support for C11 regex (e.g., GCC 4.9 meets the minimum feature set required to build the package)
  • CMake build system
  • Currently the framework can be built on Linux only

Dependencies

  • MNMLSTC Core - a C11 library that adds several library features that are to be included in C14 and beyond
  • libfcgi
  • libfcgi++
  • libxml2
  • libcrypto
  • libuuid
  • libdl

Build

Make sure the MNMLSTC Core is available in your INCLUDE path (for example, in /usr/include or /usr/local/include). It should resolve properly the header files with prefix core/ like:

#include "core/any.hpp"

To build the project you will need to have CMake installed appropriately in your system.

Inside the project directory, execute the following commands to generate Makefiles, build, and install the Fastcgi Container:

cmake .
make
sudo make install

If your default compiler does not meet the minimum feature set required to build the package, you may specify alternative compiler as follows:

cmake -DCMAKE_CXX_COMPILER=g++-4.9 .
make
sudo make install

In default configuration, the package will be installed with prefix /usr. If you want to change the prefix, you may specify it as follows:

cmake -DCMAKE_INSTALL_PREFIX:PATH=/usr/local .
make
sudo make install

or

cmake -DCMAKE_INSTALL_PREFIX:PATH=~/bin/fastcgi-container .
make
make install go

For more information about CMake options see the documentation available on official CMake site.

After installation of the package in system directories /usr or /usr/local, create the necessary links and cache for use by the run-time linker:

sudo ldconfig

Docs

Examples

The following examples are provided:

  • Simple request handler
  • Simple request filter
  • Simple servlet
  • Configuration for form authentication filter
  • Simple athorization realm

Build

All examples except C++ server page are built together with the Fastcgi Container.
To re-build them separately, execute the following commands:

cd ./example/pages
make

Configuring examples to run with HTTPD server

Copy built shared libraries from example/.libs as well as configuration file example/fastcgi.conf into working directory (e.g. ~/tmp/fscgi). Modify configuration file as appropriate.

Configure available HTTPD server to connect with Fastcgi Container.

Apache 2 HTTPD

For example, to run examples via Unix socket with Apache2 HTTPD using modules mod_poxy and mod_proxy_fcgi, add the following entries into Apache configuration file:

ProxyPass /myapp/ unix:///tmp/fastcgi3-container-example.sock|fcgi://localhost/

To run examples via tcp socket, add the following entries into Apache configuration file:

ProxyPass /myapp/ fcgi://localhost:8080/

If you want to run examples in clustered environment, enable the Apache module mod_proxy_balancer and add the following entries into Apache configuration:

ProxyPass /myapp/ balancer://myappcluster/ stickysession=JSESSIONID|jsessionid
<Proxy balancer://myappcluster/>
	BalancerMember fcgi://localhost:8080
	BalancerMember fcgi://localhost:8081
</Proxy>

Note that you will need to configure two or more instances of Fastcgi Container either on different hosts or on the same host with different ports.

For more information, see Apache documentation:

NGINX

To run examples via Unix socket with NGINX server using module ngx_http_fastcgi_module, add the following entries into NGINX configuration file:

location /myapp/ {
	fastcgi_pass unix:/tmp/fastcgi3-container-example.sock;
	include      fastcgi_params;
}

You may also run examples in clustered environment.

For more information, see NGINX doumentation:

Running examples

Start HTTPD server as appropriate (e.g. sudo /etc/init.d/apache start).

Start Fastcgi Container with exmple configuration (here the container is started under the current user; in production start it as appropriate):

cd ~/tmp/fscgi
fastcgi3-domain --config=./fastcgi.conf

Open any web browser and type the following address in URL bar:

http://127.0.0.1/myapp/servlet

or

http://127.0.0.1/myapp/myapp/test?a1=22