Asterisk PBX Part 2
Software Selection.
One of the main reasons businesses get into trouble from a technical perspective today is that they are trying to support too many disperate systems with too few staff and each time they find a new piece of software they do not take the time to see what the overall ramifications may be on the IT organization. It is much like walking into the local exchange companies (i.e. Telephone Company) of the 20th Century. Rows and rows of wires, switching equipement, and ladders to reach the wiring usually hanging 12 to 18 feet above the floor.
Our choice for testing a software-based PBX system had several initial candidates, but we quickly settled on the Asterisk PBX platform for two very fundamental reasons. These two reasons are my personal requirements for the deployment of any open-source software solution. First, the product must be mature. It must have faced a number of years of development, testing, and have a strong deployment throughout the country if not the world. Second, the product must have a commercially viable business supporting it. While I strongly believe that open-source is the only way to get some of the best minds in the world to work together to make a product the best it can be, the product is worthless to an enterprise if nobody makes any money from it. Digium and Mark Spencer, the creators of Asterisk have been around for over 10 years and Asterisk has been deployed over 2,000,000 times around the world.
We acquired a T1/E1 PCI-Express card from Digium and installed it in a IBM 2U server and installed Ubuntu Linux 10.04 LTS. Over the course of the last year I have been migrating all Linux platforms to Ubuntu so that we could standardize on one single distribution for lifecycle management. This in itself has been challenging, but that is a discussion for a different time. What was interesting was that Digium develops Asterisk on the CentOS Linux platform. CentOS being a “flavor” of Red Hat Linux, I was familiar with the distribution but was redisent to incorporating this new distribution in the company. We would see how it went.
Reading over and analyzing the vast number of Asterisk PBX distributions that now exist, my original thought was to avoid a particular distribution as it would lock the business in to a version of Asterisk PBX that may be outdated or may be too near the bleeding edge. With a source build the IT group would be able to control the version used and upgrade at their discretion. We began with Asterisk 1.8 source as provided by Digim.
Cary H. 
