BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
There are many business and technical challenges IT organizations face with today's globally distributed workforce. Enterprise WAN optimization technology addresses these challenges by increasing productivity, reducing costs, accelerating global applications and enabling collaboration. However, the pressure is higher than ever to deliver seamless services to users, irrespective of connectivity type. A common misconception is that WAN performance is always a bandwidth issue; however, adding additional bandwidth is not always the answer. The enterprise must meet the challenge of unavoidable WAN characteristics, such as limited bandwidth and network delay.
Here are five reasons why WAN optimization technology is an essential component of your enterprise strategy and how your organization may benefit from existing enterprise WAN optimization techniques, such as content delivery networking (CDN), Transmission Control Protocol (TCP) acceleration and Quality of Service (QoS).
1. TCP optimization for when TCP talks too much
Chatty applications require acknowledgements for each block of transmitted data. Each send-and-receive signal adds up to a significant delay over the distance of the wide area network, but is negligible over a local area network; this delay is called WAN latency. The application performance and user experience is dependent on WAN latency, and adding bandwidth will not always help; stopping the chatty acknowledgements before they're sent across the enterprise WAN is the key to improving performance.
To further compound the issue of chatty applications, other issues exist, including the slow-start mechanism, where TCP establishes available resources before transmission begins and packets are sent back and forth to establish conditions, adding delay. Application optimization techniques assist by locally acknowledging the traffic at Layer 7, which makes the application believe that the WAN optimization device is the server and therefore eliminates the latency restriction.
2. WAN Quality of Service for when applications don't know how to prioritize
Quality of Service has been around for some time; there's nothing new or revolutionary about it, but QoS remains an important aspect of application optimization. Service providers are adding new features to their QoS policies, such as dynamic bandwidth allocation and a more granular set of QoS classifications. While implementing a QoS policy across sites will not get around fundamental issues such as chatty applications, it's a great way to share limited resources. QoS is typically associated with a routing device, but manufacturers now offer the ability to prioritize, cache, route and implement QoS in the same device with high-availability options -- all of which provide a standardized approach and single point of contact. QoS enables you to take application types and provide a level of bandwidth and associated treatment. For example, QoS will ensure that delay-sensitive applications, such as voice, are given the best available network resources to avoid latency and jitter (a variation of time between each packet).
3. Content caching to reduce bandwidth when data is repetitive
Many organizations will send out the same data with minor changes, resulting in unnecessary bandwidth usage. An architectural firm, for example, will send out the same design files with minor changes on a regular basis. Whether you're replicating data for disaster recovery or emailing out large files, the continual repetition is wasted bandwidth -- which will have a negative impact on other applications and be costly -- unless you have plenty of headroom.
While QoS preserves the quality of key applications, content caching can reduce bandwidth by only sending the changes you've made, rather than transmitting the entire file. Regardless of bandwidth, if you're only sending out the changes rather than the whole file, costs are kept down, negative impacts on applications are minimized and the user experiences faster performance.
4. Content Delivery Networking for organizations using Web applications
More on enterprise WAN
Cloud computing performance improves with WAN optimization technology
Trends driving enterprise adoption in the WAN optimization market
Application visibility and control needed for new WAN optimization solutions
CDN helps organizations improve access to Web-based applications and content, such as images, Java, files and video. Without a content delivery network, the user is sent directly to the resources required when requesting content over the Internet, regardless of where the resources are situated. CDN replicates your Internet-based resources and redistributes them to other CDN providers' data centers around the globe. CDN providers reduce the requirement on your servers because CDN edges will only take updated content from your servers and cache content that does not change. By relying on a content delivery network, less bandwidth is required to access the content. If you're a remote user with limited bandwidth, this may greatly improve network performance.
CDN also serves up content based on a user's location. For example, a user located in the United States trying to access content in China would access local U.S. resources through a content delivery network, which results in less latency and improved performance.
Another advantage of CDN is high uptime. Since the resource exists within multiple locations, when a primary location is down, the user may be redirected to the next CDN content server. However, if your data is constantly changing, a content delivery network will not help, because updated content will constantly be requested from your servers.
5. WAN Optimization as a Service for apps in the cloud
The WAN is the fabric that connects our enterprise office locations to the cloud. Without ensuring good user performance, the cloud value proposition quickly deteriorates. To address this issue, some telecoms and WAN optimization providers are offering a cloud-based variant of WAN optimization. An enterprise should consider its own particular requirements to determine whether or not a cloud-based solution would be suitable. As an example, sites with low bandwidth and high latency would continue to require some form of on-premises hardware, since the bottleneck continues to be the lack of bandwidth and poor performance between a site and the cloud. This said, cloud-based WAN optimization providers are offering on-premises devices which are deployed within sites, where required. As the WAN marketplace embraces Ethernet with higher bandwidth possibilities and lower latency, WAN Optimization as a Service becomes increasingly attractive.
Enterprise WAN optimization continues to evolve
Without application optimization, a number of parameters must be in place in order to guarantee high performance and good business results. This is rarely the case with all sites connecting enterprise office locations. The result? Enterprise WAN optimization technology will continue to evolve and overcome less-than-optimal applications surrounding WAN pain points.
About the author
Robert Sturt is an Multiprotocol Label Swithing (MPLS) WAN procurement expert. He works for a UK-based company that focuses on assisting organizations make better decisions when buying MPLS WAN and Session Initiation Protocol solutions. As a WAN specialist, Sturt aids IT professionals in selecting an appropriate network service provider for their organizations. He also upholds procurement best practices and fosters due diligence.