At the recent Interop conference, Jim moderated a highly interactive panel that was entitled "Is Routing Undergoing a Mid-Life Crisis?" This is the fourth newsletter in a series that is based in part on that panel. This newsletter will continue the discussion of routing and the identification of what if anything is driving the need for IT organizations to rethink their approach to routing

Part of the challenge facing IT organizations as they rethink their approach to routing is that they need to ensure that the routers that they deploy today can support the current requirements and the current enabling technologies. IT organizations also have to ensure that these routers have the flexibility to provide support for the requirements and the enabling technologies that will emerge over the next several years. This is highly complex because it is impossible to predict with certainty the specific new requirements and technologies that will impact a particular IT organization in the future. It is, however, possible to state with certainty that there will be new requirements and new technologies and that the organization's routing infrastructure will have to support those requirements and technologies.

One trend that is impacting virtually all IT organizations is that the number of different Web-based applications traversing the enterprise network is growing rapidly because of the webification of enterprise applications such as ERP and CRM, plus the utilization of emerging Web-based application architectures such as SOA, SaaS, cloud computing and mashups. This trend increases the importance of identifying which Web applications are business critical so that the IT organization can provide preferential treatment to these applications vs. the more mundane or recreational applications that are also Web-based; for example, Internet radio. Routers that can base QoS scheduling and forwarding behavior on deep-packet inspection (DPI) will be able to parse application headers allowing all critical business applications, including VoIP and videoconferencing, to receive preferential treatment and enabling recreational or unwanted application traffic to be either eliminated or rate limited.

We believe that implementing router-based QoS at key points of aggregation within the network may offer an attractive alternative to managing an end-to-end QoS scheme involving numerous client end systems. We believe this because the vast majority of IT organizations that we work with have already implemented QoS. However, because of the complexity that is associated with managing an end-to-end QoS scheme involving numerous client end systems, only a minority of them can effectively manage their QoS implementation.

Steve Taylor is president of Distributed Networking Associates and publisher/editor-in-chief of Webtorials. Jim Metzler is vice president of Ashton, Metzler & Associates.