Summary

This documentation is to serve for anyone looking to integrate an enterprise XMPP/Jabber solution into both their environment and Drupal installation which would then be able to facilitate a complete federated chat solution accessible by multiple clients (Mac, Windows, Linux and Mobile). It is assumed that an environment is already in place for authenticating users and supporting servers. 

This step-by-step walk through will be using Openfire as the XMPP server, Active Directory as the LDAP repository and an Apache server to run Drupal. All installations aside from Active Directory will be running on Ubuntu 8.04 LTS.

Before we begin, I want to cover why you may consider implementing this solution. The reasoning of why I initially came up with this was to promote communication in a corporate environment. I needed a compliant and enterprise solution to provide chatting to users, but at a low cost to deploy. Using services from Google or other providers left me without control of the traffic. Having something local was required, but the solution needed to provide the ability to plug into other chatting implementations if need be.

XMPP/Jabber provides for the ability to bridge multiple chat implementations and runs using the open standard. This, among other reasons pushed me to using XMPP internally. I could host my own chatting solution (attach monitoring, control client usage, view metrics, etc.) and still allow for communication to other chat clients such as Google, Yahoo or AIM. Using XMPP also meant that I could use a wide variety of clients (web, javascript, thick, and mobile) without picking just one. 

Having picked the chatting solution, I needed to integrate that into an existing Drupal installation. Drupal was perfect, not only because the popularity of the framework, but because it too promoted collaboration and communication (through blogs, polls, and pages). What better way to extend that functionality then to add real-time chatting capabilities. Searching online did not reveal any clear and concise documentation for doing such a task, so this is why you are now reading this.