Trending studies almost always tell us something we didn't expect. Most recently, we got back the results of a survey dealing with wide area networks and their use and management. The last time we ran this survey was in late 2008, so we expected a steady progression toward buying more and higher-capacity WAN links. We didn't find that. In fact, we found a slight fall in the number of very high capacity connections such as OC-12 and bonded OC-3 lines.
To a degree, this mimics other areas where we see consumer demands outpacing business demands, but it also speaks to the reality that throwing bandwidth at most WAN performance issues doesn't do much good. App performance over the WAN is more likely stifled by latency and related problems. But all this isn't to say that the use of WANs isn't evolving. In fact, several trends we often discuss are validated by the results of this survey.
The first and most obvious trend is the increasing importance of Web-based applications. In terms of applications that must be optimized over the WAN, Web apps showed a 13-point jump, while Exchange (No. 2) showed only a two-point increase. It's also clear that enterprises are working to consolidate both data centers and other IT functions like backups. Because we went from asking about eight reasons for deploying WAN optimization to 12 options, all results saw lower response rates this year than in 2008--except for data center consolidation, which saw a six-point uptick. Elsewhere, WAN-based backups saw a substantial jump from 2008, as did database updates.
My take on these numbers is that CIOs are continuing to close satellite data centers and recentralize IT functions. While we didn't ask about it here, you can bet that two main drivers are cost control and improved auditing and procedural adherence. Not terribly interesting stuff, but it's what the business--and regulators--demand. Concern for database updates over the WAN is up, too, indicating the users outside of the main office are increasingly using enterprise applications. Along with that, we're seeing an uptick in real-time data like VoIP and, to a degree, videoconferencing.
While current WAN links may be able to handle the load, that won't last long, particularly if users start to embrace videoconferencing, and if functions such as centrally managed backups over the WAN aren't offset with commensurate data center consolidations.
My sense is that organizations have had to find ways to live with the WAN capabilities they have, given budget pressures of the past two years. Sooner or later, something must give, and most organizations will have to look at both adding WAN optimization technologies (if they haven't already) and adding bandwidth. For both of these considerations, the presence of real time, near real time, and legacy traffic on your networks warrants a fresh look at the options available.
In terms of adding more bandwidth, MPLS and VPLS overlay technologies should be seriously considered. VPLS in particular--while still in the standardization process--can provide the sort of operational flexibility that seems to be required by many businesses today. For the WAN optimization question, perhaps the biggest task is understanding application-specific capabilities now offered by vendors and how those will affect network performance. Do your homework! Simply buying more of what you bought three years ago is not the way to go.
Art Wittmann is director of InformationWeek Analytics, a portfolio of decision-support tools and analyst reports. Write to him at [email protected].
To find out more about Art Wittmann, please visit his page.
Register to see all reports at InformationWeek Analytics.
