Blurring the Lines in Networks

Once the network was considered a separate and distinct part of the IT infrastructure, but as more and more Web-based applications are produced, the lines of distinction between the application, physical hardware, and network become blurred.

Networks - Blurring the Line

As data centers become more centralized so that management costs can be reduced, and synergies extracted from the physical hardware, then the network suddenly becomes an extension of the data center, with the WAN being central to its use. When this is combined with the increasing demands being placed upon the network, and the diversity of uses that it is now supporting - such as VoIP, streaming video, and continuous data protection, for example - it is not surprising to hear that many in the industry are looking at the next generation of WAN optimization technologies as a more holistic solution than those currently deployed.

Currently, WAN optimization is provided by three separate approaches, all of which solve part of the problem, but do not address the whole issue of how to optimize the WAN. How do we know that it is running any better than it was previously for all users?

However, another complication for WAN optimization is the fact that application development protocols such as CiFS and MAPI were all designed for LAN and not WAN traffic. Even the newer protocols like HTTP, XML, and SOAP are not designed for the WAN; these new development protocols are focused on ease of development, and not designed to ease the pressure on the network.

Therefore, the challenge is to design a method that can measure the end-to-end response time from client to server so that pre- and post-optimization can be seen. It is only when WAN optimization takes this complete perspective that we will witness the convergence of networks and data centers into a single entity.

BUTLER GROUP OPINION

Butler Group considers that WAN optimization is a rapidly evolving technology, which is currently made-up of many different point or segment-based solutions. We believe that as users become more sophisticated, and demand increasing levels of evidence to support the claim that the network is delivering more performance without increasing allocated resources, then we will witness a new breed of optimization technology.