SD-WAN to Cloud Transition and Autonomous Vehicles
July 31, 2018
According to 451 Research, there are several key factors behind the recent emergence of autonomous vehicles (AVs): a combination of increasing computing capabilities (Moore’s law), accessibility of data, meaningful and actionable insights through analytics of this data, and the advent of AI, machine learning and artificial neural networks. Expertise and technology in these fields have historically been outside the core competencies of traditional automotive players (OEMs, suppliers). This vacuum has attracted the likes of new and old IT hardware and software providers to find a role in the future connected car and AV. 451 says that the largest impending innovators will come from technology providers, currently operating in a closed-house for its AV development, such as Google.
Google (Waymo) has been at the forefront of AVs, having recorded millions of miles of AV driving data, and billions more for simulated miles on its virtual AV driving simulator. AVs are essentially mobile network nodes to the Google Cloud that periodically sync for data uploads and downloads.
In essence, it’s all about visibility to the way data are captured and analyzed to optimize AV traffic patterns. A similar transformation is happening today in corporate networks. Just as the Cloud changed the economics of enterprise data centers, SD-WAN is changing the economics of long distance networks and distributed workloads. Cloud and SD-WAN have positive synergies to benefit from. In addition to cost savings, secure-Cloud SD-WAN offers virtualization, agility, automation, on-demand provisioning, elasticity of deployment and scale, and highly granular packet-level data and analytics.
For decades, enterprises have relied on private networks for moving mission-critical applications between their branch offices and data centers via single source private leased lines and MPLS circuits. Evolutionary change is occurring in which enterprise demands for greater security, bandwidth resource utilization and quality of experience are satisfied before moving application workloads to any number of Cloud services.
Technology Advances Require Adjusting to New Paradigms
Just as autonomous driven vehicles allow for the freedom to transport any person or object securely and safely, SD-WAN frees up enterprises to mix and match service providers and circuit types to best align their applications. This dynamic has given enterprises freedom to gain the reliability, cost efficiencies and flexibility they’ve never had before.
SD-WAN brings together diverse circuits to increase bandwidth and agility, while lowering costs and complexity. This is an upside-down equation for traditional enterprise networks.
SaaS, IaaS, PaaS, UCaaS and other Cloud services have swung WAN traffic workloads from the data center to the Cloud. According to the IDG – Enterprise’s Cloud Computing Executive summary, Cloud is used by 70% of U.S. organizations. This utilization includes private, public, hybrid or a mix of Cloud services, which is exacerbating congestion at the last-mile remote or branch location wherein Global IP traffic is tripling year over year, according to Cisco’s Zettabyte Index.
WANs Shouldn’t be a Roadblock for Accessing Clouds
Enterprises with traditional network architectures struggle with costly and insufficient bandwidth utilization, and security fears have caused them to inefficiently backhaul Internet-based Cloud traffic through their central data center. Traditional MPLS-WAN architectures weren’t built to support Cloud apps. These networks were designed for fixed site-to-site connections, where applications reside within the corporate data center. They are rigid and complex, and require lengthy schedules when changes are made. Legacy networks compensate for this difference, by making Cloud applications crisscross unnecessary hops, using up valuable bandwidth, and increasing packet loss and latency.
The world’s craving for more and more data grows every day. To support this hunger, Gartner estimates enterprises will need 20-50% more bandwidth each year through 2019. The high costs and long deployments of traditional networks are causing enterprises to refocus their networks to include a mix of circuits, from MPLS, DIA Internet and broadband Internet, to wireless and satellite connectivity, all seamlessly managed by the SD-WAN.
Those evaluating SD-WAN should be aware that some are better-suited to support multi-Cloud environments than others. Most SD-WAN solutions address the costs and rigidity associated with traditional WAN architectures. However, not all are optimized to support new app delivery models, where business consumes IT resources and services that are appropriately balanced with costs and technical and business efficiencies. Another way to say this is, some SD-WANs are purpose-built to support the Cloud, while others are best-suited for aggregating multiple circuits and executing last-mile performance optimizations.
The key differentiation is how seamlessly the SD-WAN directly connects users and IoT devices to Cloud services, and how business policies are defined and performed using proactive traffic steering based on application types and where they reside. This approach will inherently optimize network reliability and performance, and ensure security for every application, and for any Cloud service.
Obviously, security is a big concern when mission-critical applications travel over public Internet circuits. Again, this is where network administrators who evaluate SD-WAN need to diligently assess, via a third-party benchmark such as NSS Labs, what kind of integrated security, if any, was originally designed, tested and deployed natively within the core SD-WAN solution.