Second, latency due to distance must be addressed. This is often handled by various protocol acceleration techniques, which minimise the chattiness of specific protocols, like TCP and CIFS. By increasing window sizes and performing selective acknowledgements, for example, WAN optimisation extends the distance by which people can communicate and access data in remote locations.
The trouble with 'packet loss'
Finally, as more enterprises look to share WANs such as MPLS and the Internet to lower network costs, a new problem arises: packet loss. These shared networks will always experience congestion, resulting in packets being dropped or delivered out of order. This can have a detrimental effect on real-time traffic, like voice, video and desktop virtualisation. It also cripples traffic that requires high-sustained data throughput rates, like data replication.
WAN optimisation addresses these challenges by fixing dropped and out of order packets in real-time. These techniques avoid packet re-transmissions, improve the quality of real-time traffic and ensure data throughput rates remain high-level.
By dramatically improving WAN efficiency, IT organisations are able to improve end-user productivity, enjoy the flexibility of increasing geographical distances between offices, and keep pace with growing end user demands—all without requiring costly increases to WAN bandwidth.
When deployed correctly, the benefits of WAN optimisation are extremely tangible—from improved data transfer times to significant cost savings. By overcoming the common performance challenges that so many organisations face, WAN optimisation ultimately helps companies abide by compliance demands and enables them to take full advantage of their virtual applications. This provides a robust and rapid return on investment for any organisation.
Alan Leong is Vice President, Asia Pacific, at Silver Peak.
Sign up for Computerworld eNewsletters.