Cyclone3 Framework

developed since 2002, released under GPLv2 in 2006

Cyclone3 is extremely flexible and mature open source framework designed to develop content management systems, custom intranet applications, CLI and async job applications writen in Perl.

Basic features

• multi-engine

  framework supports multiple engines for different kind of processes: generating webpages, async jobs, cli commands, ftp server, or any custom protocol ...

• multi content-type CMS

  publish engine generates content in XHTML, HTML5, SVG, XML, JSON, RPC/SOAP services, ...

• multi-domain

  one framework installation, unlimited number of domains and services.

• multi-server

  developed for HA cluster installations in master-master mode

Used technologies

• Perl & FastCGI
• MySQL/Percona
• RedisDB
• RabbitMQ
• ElasticSearch
• Template Toolkit 2

Quick Installation

We are typically using Debian or Ubuntu for Cyclone3 Framework, so these are our two cents for Debian administrators. This installation process can take 15 minutes (without optimization).

Note: If any problem occurs, don't hesitate to ask us at open@comsultia.com.

This is just basic setup, but can be enhanced to get full Cyclone3 cluster.

Prerequisites

• Linux operating system (mostly tested under Ubuntu Server LTS 14.04 and 16.04)
• git
• Perl >= 5.12
• MySQL >=5.5
• Apache2
• root access (no root access? forget about using Cyclone3)

Getting Cyclone3

Prepare environment

Create cyclone3 user

groupadd cyclone3;useradd cyclone3 -g cyclone3 -G www-data,crontab,users -s /bin/bash -m

Add Cyclone3 binaries location to PATH and export CYCLONE3PATH. Edit /etc/profile file, or add this content into /etc/profile.d/cyclone3.sh file

CYCLONE3PATH="/srv/Cyclone3"
export CYCLONE3PATH
PATH="$PATH:$CYCLONE3PATH/.bin"

Apply environment variables

source /etc/profile

Download Cyclone3

Get Cyclone3 Framework source codes and binaries from git to directory /srv/Cyclone3 (you can use different directory, but this is default and optimal setup). Don't create /srv/Cyclone3 as symlink.

You can clone from http://bit.comsultia.com/scm/cyc/framework.git repository (Comsultia) or https://github.com/rfordinal/cyclone3-framework.git (github)

Please, don't download as zip archive, the upgrade process after installation is provided using git too.

git clone http://bit.comsultia.com/scm/cyc/framework.git /srv/Cyclone3
chmod 770 /srv/Cyclone3
chown www-data:www-data /srv/Cyclone3

Install all dependencies

Required libraries

apt-get install build-essential libinline-perl libmime-perl libdatetime-perl \
libxml-generator-perl libxml-xpath-perl libxml-simple-perl libsoap-lite-perl \
libnet-smtpauth-perl libstring-crc32-perl libfcgi-perl libcgi-fast-perl \
libparallel-forkmanager-perl libfile-type-perl \
libjson-any-perl perlmagick libtime-modules-perl libxml-libxml-perl \
libjson-perl libproc-processtable-perl libtie-ixhash-perl libmoosex-getopt-perl \
libanyevent-perl

Perl libraries not available in Debian/Ubuntu as .deb packages

cpan CPAN
cpan Template::Stash::XS
cpan Digest::MurmurHash
cpan Digest::SHA1
cpan AnyEvent::ForkManager
cpan Sys::Info
cpan Net::RabbitFoot

Install and setup database

MySQL

MySQL >= 5.5 is the main database for Cyclone3, but connection to another type of databases can be configured too.

We recommend Percona Server (http://www.percona.com/)

sudo apt-key adv --keyserver keys.gnupg.net --recv-keys 1C4CBDCDCD2EFD2A

Add this to /etc/apt/sources.list, replacing VERSION with the name of your distribution:

deb http://repo.percona.com/apt VERSION main

sudo apt-get update && sudo apt-get install percona-server-server percona-server-client

Log in to MySQL and create Cyclone3 user (To exit MySQL type exit)

mysql -h localhost -u root -pPassworYouSetEarlier

GRANT ALL PRIVILEGES ON *.* TO 'Cyclone3'@'localhost' WITH GRANT OPTION \
MAX_QUERIES_PER_HOUR 0 MAX_CONNECTIONS_PER_HOUR 0 MAX_UPDATES_PER_HOUR 0;

mysql -h localhost -u Cyclone3 < /srv/Cyclone3/_data/TOM.sql

RedisDB

apt-get install redis-server
cpan RedisDB

Optimize redis.conf

# much much faster
unixsocket /var/run/redis/redis.sock
unixsocketperm 777
# save io :)
save 86400 1
save 7200  1000000
save 3600  2000000
# save CPU
rdbcompression no

Setup Cyclone3

Copy configuration template file to destination

cp /srv/Cyclone3/_config/TOM.conf.tmpl /srv/Cyclone3/_config/TOM.conf

This is minimal working configuration example:

#!/usr/bin/perl
package TOM;
use strict;

$TOM::domain='nameofmycluster'; # "domain" of cyclone3 cluster nodes
$TOM::contact{'from'}='cyclone3@'.$TOM::hostname;
$TOM::contact{'_'}='cyclone3@'.$TOM::hostname; # default email to send notifications about health

# sending email over ...
$TOM::smtp_host='localhost';

# default databases
$TOM::DB{'main'}= {
	host	=>"localhost",
	user	=>"Cyclone3",
	pass	=>"",
	sql => [
		"SET NAMES 'utf8'",
		"SET CHARACTER SET 'utf8'",
		"SET character_set_connection='utf8'",
		"SET character_set_database='utf8'",
	]
};
$TOM::DB{'stats'}=$TOM::DB{'main'};
$TOM::DB{'sys'}=$TOM::DB{'main'};

# apache2 user
$TOM::user_www="www-data";

#$Ext::Redis::host='/var/run/redis/redis.sock';
$Ext::Redis::host='localhost:6379';

1;# don't remove me!

Read more about all configuration options in TOM.conf

Check installed MySQL database scheme

tom3-chtables

Check files permissions

tom3-chfiles

Setup Webserver

Install apache2 and mod_fcgid

apt-get install apache2
apt-get install libapache2-mod-fcgid

Add www-data user to cyclone3 group

usermod -a -G cyclone3 www-data

Copy virtualhost default template

cp /srv/Cyclone3/_config/httpd.virtual.conf.tmpl /srv/Cyclone3/_config/httpd.virtual.conf

Symlink configuration files

ln -s /srv/Cyclone3/.core/_config/httpd.conf /etc/apache2/conf-enabled/00-cyclone3.conf
ln -s /srv/Cyclone3/_config/httpd.virtual.conf /etc/apache2/sites-enabled/01-cyclone3-virtual.conf

Enable mod_expires and mod_rewrite

cd /etc/apache2/mods-enabled
ln -s ../mods-available/expires.load .
ln -s ../mods-available/rewrite.load .

Restart apache2

service apache2 restart

Now is everything configured properly, the next step is configuration/installation of first domain service (virtualhost)

Optimal setup

• TODO: RabbitMQ and job.worker[d]
• TODO: ElasticSearch
• TODO: rlog ElasticSearch

Advanced setup

• TODO: Cluster setup