Add commentsNov 18, 2009
Virtualizing Voice
We read with great interest this article in Virtualization Journal which asks the question why aren’t we using virtualization for unified communications, specifically for voice applications? This is really a great question. After all, as Certeon points out in the article, if we were to move from the current voice delivery hardware-based solutions to a virtual software appliance voice delivery model, some of the savings could include:
- Acquisition costs: 37% savings
- Operations costs: 33% savings
- Maintenance and Support costs: 36% savings
A no brainer, right? So, why aren’t we delivering voice via virtual software appliances today – well, as the article points out, it’s really hard.
Let’s take a look at the authors definition of unified communications for voice applications. Unified communications is the convergence of communication networks, collaborative and business applications plus devices. It combines voice, messaging, mobility, conferencing and collaboration applications. Some of the promised business benefits of unified communications include: lower communication costs, improved employee efficiency and productivity, enhanced responsiveness to customers, suppliers, and partners, improved IT management, and lower total cost of ownership (TCO).
The business benefits of unified communications voice delivery through virtualization would appear to support a virtualized software environment, but there is a challenge, that challenge is latency. While latency is okay for certain applications and in areas like trading for financial services is top of mind, latency is unacceptable for voice applications. So what needs to happen to minimize latency and deliver unified communications voice delivery through virtualization? First, organizations need to bring voice into the data center and virtualize it using software appliances and test for usability. As the author points out, this testing will “break down the barriers between the telecom closet and the data center.” IT managers can then use virtualization to “minimize voice management, consolidate voice and non-voice applications on a single server.”
This is a very interesting application of software virtualization and one that could show huge savings in the data center. What, if anything, are you doing with voice virtualization? Are you considering it? Have you already done it? Are you looking at hardware or software as the solution?

