Why is capacity planning for SD-WANs important? It helps ensure that service provider links are being utilized to their full potential, thus providing adequate bandwidth and quality of experience for an organization’s critical applications.
Gartner forecasts that by the end of 2021, more than 65 percent of enterprises will have deployed SD-WAN technology. By nature, SD-WANs are complex and often part of a multi-vendor, hybrid environment that includes legacy infrastructure, data centers, cloud components, various service providers and a vast assortment of routers, switches and endpoints distributed across remote sites. As SD-WANs continue to become a mainstay among modern businesses, those failing to properly assess network requirements, and provision resources effectively, will face costly inefficiencies and service issues.
The value of capacity planning for SD-WANs
Businesses are implementing SD-WAN technology for many reasons. The main draw is enabling the level of agility and flexibility needed to intelligently route traffic across complex, distributed environments depending on the needs and priority of each application. In these environments, capacity planning is ultimately the lynchpin for sustained performance and reliability.
This piece of the puzzle is critical during the planning and design phases of new SD-WAN deployments in order to understand existing and projected bandwidth usage across various sites, and properly adjust capacity. But it doesn’t stop there. Ongoing capacity planning plays a key role in the management and optimization of any SD-WAN as it evolves over time.
Properly provisioned SD-WANs can provide reliable support for critical line of business requirements regardless of application origin or type. They can eliminate the risk of congestion and interrupted service, resulting in positive end-user experiences. And they can root out network resource inefficiencies, which eliminates unnecessary capital spend and improves the bottom line.
The rapid adoption of SD-WAN technology carries with it inherent complexities that need to be simplified and managed strategically. This all starts with capacity planning. Let’s examine a few major points to consider when planning for network capacity and how this process can streamline SD-WAN deployment and management.
Analyzing interface capacity – What’s the best way to determine if QoS policies are being applied as intended and if key applications have been allocated with enough bandwidth? By analyzing bandwidth percentage changes over a period of time for specific sites, service providers, etc. In doing so, you can establish an understanding of normal usage, analyze which interfaces might be over- or under-subscribed, and cut costs through adjustments to capacity allotment. One of the major benefits of SD-WAN is that users can easily change service providers, due to the “over-the-top” nature of the technology. As a result, lower cost service providers that are enabling a good quality of experience can be utilized more often throughout the enterprise to capture cost benefits.
Establishing accurate utilization metrics – During the planning and baselining process, some admins might be misled by usage measured over certain time periods. For instance, bandwidth utilization is typically far higher during business hours than at night or in the early morning. If usage during business hours and off-peak hours are taken together as an average utilization bandwidth, the data from non-business hours will skew the bandwidth requirement data, causing sub-optimal performance during peak business hours. You need a complete picture of bandwidth usage and capacity requirements, visualized in a way that intelligently delineates the data between peak hours and non-peak hours.
SD-WAN not only allows users to better utilize their WAN circuits, but it also provides key metrics on the performance of the links as well. Here, ongoing capacity planning offers deeper insights into the performance metrics of the link, including loss, latency and jitter.
Tracking application usage – Modern SD-WAN networks are highly application-aware and can route traffic intelligently based on individual application needs. SD-WAN vendors are able to accurately identify thousands of applications by means of deep packet inspection. Therefore, capacity planning should provide users with insight into their critical application usage across all WAN circuits. For example, if a customer identifies that the majority of their critical traffic, like voice and video, is taking a lower cost internet circuit with high quality, as opposed to the more expensive MPLS provider, adjustments can be made to reduce MPLS cost. Various sites may also be good candidates to eliminate MPLS altogether and incorporate multiple business internet links. Tracking critical application usage across the network is very important when trying to understand capacity planning. This needs to be done with true DPI-based application information, as opposed to legacy port/protocol-based methods.
Where to go from here
As the pace of SD-WAN adoption quickens, capacity planning will become increasingly important for successful deployments. Every business needs the ability to visualize these complex networks, gather relevant bandwidth utilization data from their component parts and make well-informed capacity adjustments that improve efficiency and performance. Without the necessary data and granular insights during the planning process, it can be nearly impossible to operationalize and manage high-functioning SD-WANs. There’s no question – we’re seeing the future of networking unfold before us. If you’re considering a new SD-WAN initiative this year, be sure to prioritize preliminary and ongoing capacity planning to ensure that future is bright.