Fredrik Lindstrom, Manager CIO Advisory, KPMG
With the transition from controlling and using physical and digital assets in the Third Industrial Revolution, the Fourth Industrial Revolution now adds a biological element. While this may seem like a minor addition, combining physical, digital, and biological will change how we work, live, transport ourselves, and ultimately, it will change us as humans.
CIOs and IT organizations struggled with the rate of change during the Third Industrial Revolution, especially in the final phases with the “consumerization of IT.” Now, the rate of change will move from linear to exponential, presenting an unprecedented challenge for CIOs and IT organizations across all industries. This challenge will add to the pressure CIOs are already struggling with today—doing more in less time with fewer resources—and further drives the traditional approach to infrastructure management to obsolescence.
As change accelerates, the key issue is the misalignment between infrastructure, specifically network investment, and the shorter strategic business horizon. CIOs can easily create the plan, build and then run organizations based on a strategic business plan that spans five years with an infrastructure lifecycle of five to seven years. When the strategic horizon shrinks as speed of innovation accelerates, the CIO needs to adapt to three-year plans. In some cases, even the next quarter is unknown as new technologies are made available with new and different demands on the network.
Network Under lays and Overlays
To meet the challenge of this ever-increasing speed of change, the obsolete way of network management— with standalone devices that must be configured individually using syslog and SNMP messages—simply does not work. Nor does the traditional way of thinking in tiers of services, with Core, Aggregation and Access.
"The virtualized network provides automation and integration capabilities that tie into several different orchestration tools"
Both the physical design and the logical configuration must change to meet the new demands on the network. Networks must be faster, easier to manage, more secure, scalable, and provide analytics to constantly adapt and improve.
Solving these problems starts with flattening the architecture and driving speed efficiencies and cost savings. Ethernet fabrics provide network design where the entire datacenter can be managed as one unit, simply making the network easier to manage and significantly faster. As the demands grow and more capacity is needed, the network can scale without downtime, and simply by adding what is required, truly become a pay-as-you-go design.
The deployment of a network overlay, virtualizing the network, further increases its speed and flexibility. With a virtualization network, it is possible to automate most network tasks for building, running, and sun-setting services on-network. The virtualized network makes it extremely easy to map automation to existing processes, making the network easier to manage and maintain.
The virtualized network provides automation and integration capabilities that tie into several different orchestration tools. Orchestration enables the enterprise to deploy compute, storage, and network seamlessly as one unit. If a new web server is needed, it is automatically provisioned based on the enterprise standard. It is even possible to add specific security controls automatically—from firewalls, intrusion prevention, and advanced malware protection—that can be deployed seamlessly with the web server. These controls scale with the environment. And as additional resources are added to the compute clusters, more resources are added for the controls. This makes the virtualized security controls a great match for inspecting and controlling east-west traffic in the datacenter, even between virtual machines on the same physical host.
The Missing Piece Found - Analytics
Traditional networks powered the Third Industrial Revolution and lasted for decades, but provided very little insight into what actually happening on the network. At best, the traditional network could provide the amount of bandwidth that was consumed with the help of SNMP.
When net flow was introduced, additional visibility was provided, but this was still network-centric. What is lacking is an application-centric approach that provides a holistic view of application performance behavior, both in the virtual and physical infrastructure.
Analytics solves this challenge by providing an application-centric, end-to-end view of performance and behavior. The analytics engine can correlate issues in the virtual network layer with issues in the physical network. This simplifies capacity management and troubleshooting, driving down costs associated with downtime by proactively identifying and mitigating network faults.
An important capability of an analytics engines is an open API with an extensive partner eco-system, proving an easy integration point to other tools. With the right API and eco-system, the analytics engine can provide a single pane of glass for the virtualization team, allowing the virtualization team to view performance and troubleshooting information for all components in the virtualized and physical environment.
CIOs and IT organizations are facing unprecedented challenges scaling and managing the infrastructure environments, and the speed of change is ever-increasing. The key to conquering these challenges is utilizing the right infrastructure. With an Ethernet fabric, virtualized overlay and advanced analytics, CIOs and IT organizations can rest easy, the network is ready for the unknown of tomorrow.