Hershey’s Reese’s Peanut Butter Cups commercials from the 1970s and 1980s famously declared their candy had, “Two great tastes that taste great together.” It may be coincidence that the Internet of Things (IoT) and virtualization are evolving in the same time era of technology, but the serendipity of their co-existence cannot be dismissed. If, as Cisco (News - Alert) projects, IoT is to scale to 50 billion devices by the year 2020, it will require an agile and elastic infrastructure that only a mature virtualized cloud-based architecture can deliver. Looking at this from the opposite perspective, the virtualization of the network infrastructure may or may not see the return on investment in terms of cost and operational benefit if the demand for IoT does not develop.
Visibility and automation is key
The drive for virtualization began with cloud technologies and has evolved to include technologies like SDN and NFV. The lynchpin that holds the key to success for all of these cloud-enabled, virtualized architectures is the need for a unified and holistic management and orchestration system. This system must be able to understand the disparate components within the architecture through analytics, have the heuristics intelligence to analyze the data from a broad ecosystem perspective and orchestrate the environment through policies, as well as have the ability to enact changes automatically to adjust for shifting conditions.
The base cloud technologies are relatively mature and proven. Application delivery technologies based on application load balancing and global server load balancing (GSLB) functions enable the cloud infrastructures to be available, resilient, and scalable. Today, there is considerable debate around the value in new virtualization technologies such as SDN and NFV. It remains to be seen if they are a passing fad or something of value. SDN is still being defined and NFV is barely out the gate, with isolated proof of concepts being tested.
Meanwhile, IoT must find a way for devices with diverse functions and a myriad of connectivity models to function within the networks and Internet of today. WiFi (News - Alert), Bluetooth, and cellular wireless are just a few of the access technologies IoT devices will use to connect to the Internet. IoT must do this while scaling towards the predicted exponential growth to support billions of devices and applications. There are already wireless smoke detectors, connected cars, personal fitness aids, home security and automation systems, and kitchen gadgets including refrigerators and crockpots.
IoT presents the biggest challenge to network design that network architects have seen in a long time. To properly manage the number of devices, their harmonic connections, and bandwidth consumption, a massive network infrastructure must be put in place to manage the peak loads.
Virtualization provides the agility and elasticity for networks to deliver the resources necessary for IoT to flourish, without the need to overbuild network infrastructures. Network resources can be allocated and de-allocated as needed by the network-aware, ecosystem-aware management and orchestration system.
The system must be able to collect information from unrelated technologies and multiple vendors. Data from network components, application health metrics, DNS information, routing topologies, and many other elements must be collected. The management and orchestration system must understand the inter-relationships between these different data points relative to the specific architecture and configuration of the virtualized infrastructure.
Once the data is collected, an intelligence needs to exist that understands the inter-relationships between these data points and how it relates to the delivery of the business applications and, in turn, understand how the application is impacted by the real-time analytical data. Individual customer and application-based policies have to be programmed into the system to provide a functional understanding that the heuristics engine can leverage to present meaningful and holistic information.
Once the intelligent management and orchestration system understands the architecture and behavior of the application delivery infrastructure, it can make recommendations and provide insight to enact changes to the current environment to adjust the cloud ecosystem to optimize the application delivery based on different application service level assurance (SLA) levels. Not all applications and functions are created equal. The HVAC automation system has different resilience and performance requirements than the assembly line automation infrastructure, which, in turn, has different requirements than the personal health and monitoring devices.
Ultimately, it is essential to automate the processes of the management and orchestration system because it is not feasible to have the analytics, detection, and proscriptive changes manually driven for each application’s unique SLA requirements 24/7/365. Removing the human element removes the chance for human error and reduces the operational needs to support the infrastructure that may have thousands of devices and applications on it, if not billions.
IoT and Virtualization – Convergent Evolution
IoT gives virtualization a clear and focused reason to exist and, in turn, the successful evolution of IoT depends on the flexible network framework that virtualization can deliver, forming a truly symbiotic relationship.
It is important to note that the IoT devices and applications exist in a multi-technology/multi-vendor environment and the virtualized cloud infrastructure will also consist of multiple technologies and vendors. The management and orchestration system must be flexible and extensible enough to incorporate these different aspects to deliver a unified and holistic view of the entire application delivery ecosystem.
There are potential benefits with either of these technologies by themselves, but together, like the Reese’s Peanut Butter Cup, they synergistically work together to push us more quickly to the next generation networks and applications.
About the Author: Frank Yue is the Director of Application Delivery Solutions for Radware. In this role, Yue is responsible for evangelizing technologies and trends around Radware’s ADC (News - Alert) solutions and products. He writes blogs, produces solution architectures, and speaks at conferences and events around the world about application networking technologies. Prior to joining Radware (News - Alert), Yue was at F5 Networks, delivering their global messaging for service providers. Yue has also covered deep packet inspection, high performance networking, and security technologies. Yue is a scuba diving instructor and background actor when he is not discussing technology.
Edited by Maurice Nagle