Network Optimization

Introduction to WAN Optimization Techniques

Ever wondered why WAN optimization is required in consolidated data centre scenarios? Or how much difference WAN optimization makes in such situations? This article gives an overview of the techniques that are applied to optimize the data that is sent across the WAN network in order to reduce the bandwidth consumed by them and also to reduce the delay associated with WAN links.

What is WAN Optimization and why is it required?

Wide Area Networking has enabled a lot of applications (both for connectivity as well as optimization). We can today use the Internet Leased Lines and MPLS VPN Connectivity for connecting different branches of an organization. These are quite cost effective when compared to IPLC or dedicated Leased Line connectivity. One of the consequences of the affordability of such higher bandwidth and connectivity is the consolidation of servers and core networking components in a single data centre. This has definitely reduced a lot of cost (when compared to individual servers that needed to be replicated at each branch, in a few cases) and enabled applications like Virtualization (Where the resources of a server are shared between multiple applications as there is always umpteen processing power that is not utilized at all points of time, for example).

But consolidating resources through Wide Area Networking also comes with its own problems. There is a delay or latency induced in the packets that has to travel all the distance. This delay might also be due to the line conditions, amount of traffic at that instant, hardware induced delays like buffer delays both at the local end and ISP end. And there are always real time applications like voice, video etc. that travels along with bandwidth intensive applications like email and ERP that might cause un-acceptable delays if proper QoS policies are not set.

So, all these factors plus a lot more creates the need to have WAN optimization processes in place. Although the WAN optimization techniques do not eliminate the delays, they reduce it to a good extant. And WAN optimization induces its own security concerns but fortunately WAN optimization techniques address security in a much broader sense due to their unique position in the enterprise network where all the packets need to pass through such devices.

WAN Optimization Techniques:

Protocol Substitution:

¤ WAN Optimization devices manage chatty protocols inside the LAN and then encapsulate and manage communications across the WAN using less chatty protocols. For example, TCP (Transmission Control Protocol) requires acknowledgements for transmissions received and includes other mechanisms to do windowing, congestion management, slow start, etc. to manage its performance characteristics which induce further delays. Such TCP packets are often repackaged inside UDP packets for WAN transport or applications are re-architected to replace TCP with UDP for WAN Optimization.

¤ Multiple long messages are consolidated in to one or more longer ones. Each protocol has its own size of packaging messages. Video packets for example are longer than http packets. If the packets are small, they need to be sent and acknowledged more often and delay sensitive packets (which are longer) might have to wait. So, it is more efficient to combine a number of such small packets and send them in one go to reduce the number and frequency of such packets to traverse the WAN Link and to enable designers to choose optimization techniques uniformly for such WAN friendly message structures.

Data Substitution/ Cashing:

WAN optimization can look for recurring strings and data elements at the application layer. If these elements are copied across the WAN Link once, and they don’t change thereafter, a couple of WAN optimization devices can index them and exchange only index values between them so that frequently sent data elements need not be sent again and again through the WAN Link. For example, if a website like the company website is accessed by many members in an organization frequently, then most of the data of that website could be stored locally in a cache and sent to the employees instead of contacting the remote web server other than transactions which involve some real time processing like authentication etc.

Data Compression/ Encryption:

A variety of compression techniques are used to compress the WAN traffic. Header and Payload compression techniques use pattern matching algorithms to identify short, frequently recurring byte patterns that can be replaced by even shorter segments to reduce the final transmitted sizes. Different techniques might be employed to compress data across the WAN link but data which are already compressed like ZIP files or H.264 video packets etc. cannot be compressed much again. The traffic that is sent in the WAN links are often encrypted for security reasons as public networks are more and more utilized these days.

You can stay up to date on the various computer networking technologies by subscribing to this blog with your email address in the sidebar box that says ‘Get email updates when new articles are published’

One Comment

  • Michael Leonard

    This is a good overview of WAN Optimization. I’m in marketing at Cisco for WAAS, our WAN Optimization solution. We are seeing a lot of uptake for WAN Optimization as customers consolidate branch office servers to the data center. In addition to the techniques mentioned here WAAS has a number of features such as transparent network integration, that preserves services on the network, and a module for the Integrated Services Router that brings down the total cost of ownership that customers find compelling.

    meet WAAS fans at:
    follow for news: