Enterprise WAN Optimization for the public cloud

The public cloud services market is projected to grow 17.5% this year according to Gartner. The fact that public cloud services have sustained such a high growth rate despite having seemingly reached ubiquity is impressive. It also reinforces the importance of WAN optimization for the public cloud.

Traditional telco-managed WAN connectivity solutions like MPLS (Multiprotocol Label Switching) simply aren’t practical for modern public cloud workloads. When it comes to public cloud, MPLS is simply too inflexible and expensive. In fact, Gartner cites the decentralized nature of cloud data traffic flows as one of the drivers of the decline of MPLS.

If traditional telco solutions aren’t the answer, where should enterprises turn for public cloud WAN optimization? The answer is communication hubs, which are a type of cloud-based network hub.

Why are they effective for modern WAN optimization? In this piece, we’ll explain.

Communications hubs 101
In a nutshell, communication hubs are cloud data centers that emphasize low latency connectivity to cloud services and other communication hubs. Physically, communications hubs consist of server racks full of switchgear and network equipment. The equipment is located in colocation data centers (a.k.a. “colos”) around the globe. Each site where a hub is located is known as a PoP or Point of Presence. The PoPs are connected using low-latency connections. The interconnected web of PoPs creates a high performance network backbone. The result is a high-speed, global, and private network of PoPs.

How communications hubs enable public cloud WAN optimization
While that description of communications hubs explains what they are at a high-level, it doesn’t explain how they can be used to improve cloud connectivity. This becomes clear when you consider that the PoPs are often in or near the same data centers as the major cloud service providers and communications hubs have peering relationships with them as well. As a result, the hubs can provide predictable low-latency connectivity to cloud apps and services ranging from Amazon AWS to Microsoft Office 365 to SalesForce and more.

Enterprises can connect their branch locations, mobile users, and corporate data centers to the nearest PoP using secure tunnels. By connecting to the geographically closest PoP, latency is reduced. Further, because communication hubs host their own security stacks, inbound and outbound traffic can be analyzed for malicious behavior. This helps reduce the need for CPE (customer premises equipment) from a security perspective.

Integrating communications hubs into the WAN, enterprises can benefit from:

• A low latency WAN backbone– When all apps and services were on-premises, enterprises demanded low latency LAN (local area network) connectivity. With the shift to public cloud, a low latency WAN backbone is needed for optimal performance. With a network of purpose-built PoPs, communications hubs provide just that.
• Improved mobile connectivity– With solutions like MPLS, the performance of cloud-based services for mobile users was often reduced due to inefficient backhauling. Dubbed the “trombone effect”, this backhauling negatively impacted performance and user experience (UX). With communications hubs, the backhauling goes away and public cloud application performance is improved.
• Enhanced performance for real-time cloud services– Real-time apps like those found in UCaaS (Unified Communications as a Service) solutions demand low latency. If you rely only on the public Internet or have to engage in inefficient backhauling, reliably achieving low latency connections to cloud service providers may be out of reach. The geographical proximity of communication hubs coupled with their peering relationships enables enhanced performance and low latency for many cloud services.

Cloud-based SD-WAN as a communication hub

One of the premier examples of a communication hub is cloud-based SD-WAN. With a network of PoPs in almost every geographical region making up a private backbone, cloud-based SD-WAN is capable of providing WAN optimization at a global scale. Premier cloud-based SD-WAN providers target having a PoP within 25 milliseconds of their customer locations, which is more than fast enough for latency sensitive services like videoconferencing and VoIP (Voice over IP). With this network of PoPs, cloud- based SD-WAN can deliver connection speeds comparable to services like AWS Direct Connect or Azure ExpressRoute without the additional costs.

With cloud-based SD-WAN, getting connected is simple. Physical locations can connect to the global cloud backbone using an IPsec tunnel from existing CPE or a zero touch SD-WAN device. Mobile users can connect using a simple mobile client. In each case the end result is a simple, efficient, and secure connection to a low latency backbone that provides optimized connectivity to public cloud services.

Beyond the global SLA-backed private backbone, the software running at PoPs further improves WAN optimization efforts and security. With cloud-based SD-WAN, features like user-aware Quality of Service (QoS), dynamic path selection, and packet duplication help ensure optimal performance. Further, built-in security features like anti-malware, secure web gateway (SGW), and Next-generation Firewall (NGFW) help ensure secure connectivity while reducing the need for CPE appliances and backhauling.

Cloud-based SD-WAN provides WAN optimization for modern public cloud workloads

With more and more frequency, the public cloud will become the primary host for enterprise workloads. At the same time, services like UCaaS will demand more low latency bandwidth than ever before. To ensure the WAN doesn’t become a bottleneck that limits performance, enterprises need a modern WAN optimization solution that can keep up. With cloud-based SD-WAN, enterprises can achieve low latency connections to public cloud workloads in a simple, secure, and scalable fashion.

Featured image by JuralMin pixabay