Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

Converged networks: No longer if, but when

Julie Sartain | July 17, 2008
Networks today must evolve into a robust, services-oriented platform that enriches emerging, innovative, composite applications and yet still manage to balance technology and business processes.

Bob Mays, director of network and communications at Villanova University, says, "Last year, the university partnered with Avaya and Extreme Networks to install an IP-enabled PBX with VoIP capabilities. Now, we can manage both the voice and data networks using one platform, which is a better utilization of staff because both the voice and data network teams can troubleshoot the same network. In addition, it positions the university to begin saving on long-distance expenses when outbound calls are going over the Internet."

Villanova is on a second-generation converged network built with a network foundation from Extreme and IP telephony solutions from Avaya. The network connects more than 3,000 IP and digital phones, providing flexibility for staff. Staff members use feature-rich IP or SIP phones as buildings are updated on a rolling technology upgrade.

Drawbacks

As with any architecture, notes Davis at Brocade, there are pros and cons. One possible con of converged networks is the increased complexity of managing multiple traffic types and the service levels for each, compared with a dedicated network model, where sharing is not a consideration. With any architecture that shares resources among multiple functions, the possibility of contention for resources exists. That's why features such as adaptive networking are vital in converged networks, because administrators can set policies and rules for services levels, end to end, through the converged network. This makes management easier while also enforcing the service levels, so users of the network experience the same level of service as if they were on a dedicated infrastructure.

"In addition," says Davis, "While some organizations could consolidate mainframe and open systems connectivity on platforms they already have deployed, they choose to keep these applications separated on dedicated network infrastructure for reasons of business rules, oversight and ownership, and/or organizational preference. Keeping the networks separate means they can maintain proven operational and organizational ownership and processes that would otherwise need to be reengineered in a new model."

"Another potential disadvantage of a converged network is the dependency on a single infrastructure," adds Heath at Extreme Networks. "This makes resiliency and reliability vital. In addition, a modular software operating system is also a key factor because it provides reliability through the separation of processes, hitless software upgrades and the simplicity of the same OS employed across the switching product portfolio, which minimizes human error."

According to Heath, less than 50ms path protection failover is a requirement on voice networks, a condition that may not have been imposed on the network before convergence was introduced. The Ethernet Automatic Protection Switching (EAPS) protocols can meet this need to keep a voice call up and active, even in the event of a fiber cut or interruption. Quality-of-service (QoS) prioritization for data, voice and video also becomes more important.

 

Previous Page  1  2  3  Next Page 

Sign up for Computerworld eNewsletters.