Author: Jim Metzler, Ashton Metzler and Associates
As an analyst, I advise Fortune-500 enterprises on network transformation. Some questions I keep getting asked are, why do Cloud applications perform poorly? What will happen when I migrate my entire workforce to Office365? And, how are some of the trailblazing enterprises tackling the networking issues tied to Cloud. In order to understand the crux of the problems, one must dig deeper into the WAN legacy.
The modern WAN got its start in 1969 with the deployment of the ARPANET which was the precursor to today’s Internet. In addition to the continued evolution of the Internet, the twenty-year period that began around 1984 saw the deployment of four distinct generations of enterprise WAN technologies:
- Time Division Multiplexing (TDM);
- Frame Relay;
- Asynchronous Transfer Mode (ATM);
- Multi-Protocol Label Switching (MPLS).
Unfortunately, there hasn’t been a major new WAN technology introduced into the market in over a decade. Enterprise networks, today, have been consolidating to two WAN services: MPLS and the Internet. Yet, enterprise applications have gone through a large-scale shift with rapid Cloud adoption. So, how can the WAN keep up when application models are changing in revolutionary ways? Lets investigate. (For in-depth understanding, you should download my 2015 WAN Architecture and Design Report).
The traditional approach to designing a branch office WAN is to have low-speed access to a service provider’s MPLS network at each branch office and to have one or more higher speed links at each data center. In this design, it is common to have some or all of a company’s Internet traffic backhauled to a data center before being handed off to the Internet. This design was widely adopted years ago when the vast majority of an organization’s traffic was intra-company data traffic that flowed between branch office and data center.
I recently co-authored a report entitled The 2015 State of the WAN Report. That report contained the results of a survey that was recently taken by over 100 IT professionals. I asked the survey respondents to indicate how much of their Internet traffic was backhauled. Half the respondents indicated that they backhaul the majority of their Internet traffic and 40% of the respondents indicated that they backhaul more than 80% of their Internet traffic.
Backhauling Internet traffic enables a network organization to have control over this traffic prior to handing it off to the Internet. However, one of the limitations of this design is that since the Internet traffic transits the MPLS link, this adds both cost and delay. One of the key trends in our industry is the continually increasing use of public cloud services. The State of the WAN Report contained the results of a question that asked the respondents to indicate which applications were driving the biggest increases in the use of the Internet. By a wide margin, the biggest driver was public cloud services.
Cloud providers strive to improve application performance with multiple options, like, global DNS, geographical distribution of data centers, direct peering, and more. But, the rigid manner in which WANs have been architected tends to nullify the benefits of these methods. The vendors that have solved these problems have done so by bringing forward a dynamic new architecture that is changing the WAN in a way that is similar to how virtualization transformed compute. This new approach is called Software-Defined WAN and I will get into the solution details in later blogs.
The growing use of public cloud services isn’t the only reason why the traditional WAN design model is under attack. In addition to the constant pressure to reduce cost, other reasons include the growing requirement to carry increasing volumes of intra- and inter-company voice and video traffic, increase security and improve application performance. In subsequent blogs I will continue the discussion of the WAN – where is it now and where is it heading?