This white paper is FREE to download and read. Please sign up to download.
As the industry moves forward to meet the enormous demand for data with video, mobile, and cloud, the core networks need to transition from 10G to 100G and beyond. Consider some of these findings from Cisco’s 2013 Visual Networking Index (VNI) report:
● Global network users will generate 3 trillion Internet video minutes per month, which translates to 6 million years of video per month, or 1.2 million video minutes every second, or more than two years’ worth of video every second.
● Wi-Fi and mobile-connected devices will generate 68 percent of Internet traffic by 2017.
● More traffic will traverse the Internet in 2017 than the total amount of traffic that traveled the Internet cumulatively from 1984 to 2012.
● The “Internet of Things” has arrived, and applications such as digital health monitors, smart meters, asset and package tracking, chips for pets and livestock, and video surveillance are driving more and more traffic. Globally, machine-to-machine (M2M) connections will grow three-fold, from 2 billion in 2012 to 6 billion by 2017. Annual global M2M IP traffic will grow 20-fold over the same period – from 197 petabytes in 2012 to 3.9 exabytes by 2017.
While this dramatic growth occurs, projections call for the cloud to account for nearly two-thirds of data center traffic and for cloud-based workloads to quadruple over traditional servers. That adds another element to the picture: changing traffic patterns. Under a cloud model, a university, for example, can build its network to handle average traffic volumes but then offload data on heavier trafficked days to a public cloud service when demand dictates, such as when it’s time to register for the next semester of classes. The cloud introduces new stresses to the network and creates new traffic patterns. Data centers traditionally hosted north-south traffic, with data traveling from the data center to the user. Now, however, traffic patterns are shifting, with information flowing from data center to data center in what’s considered an east-west route. That increases the number of interconnects and makes the inherent traffic patterns more distributed.
In addition, the difference between average and peak traffic volumes is increasing, and event-specific triggers, such as news events and/or natural disasters, prompt surges in traffic that are impossible to predict. Networks need to react faster than ever before. In the past, it was enough for service providers to identify major traffic patterns, bolster network infrastructure around those patterns, and periodically adjust as needed. Now, with so many changes to traffic patterns and the unpredictable nature of that traffic – a website suddenly becomes popular or a new mobile device is introduced – service providers need to be able to react within minutes rather than waiting weeks or months. They also need to meet increasingly stringent service-level agreements (SLAs) to retain their top customers.
