Seven requirements for next-generation edge networks

Enterprises that have embarked on digital transformation initiatives quickly find that their existing networks cannot deliver the experiences that today’s users demand, nor are they flexible enough to meet changing needs. These limitations are exacerbated at the edge, where the Internet of Things and workforce mobility are driving exponential growth in the number of devices connected to enterprise networks. The amount of data at the edge is also growing by orders of magnitude.

Gartner predicts that by 2025, 75% of enterprise-generated data will be created and processed at the edge, outside of traditional centralized data centers or clouds. This makes it more difficult for IT departments to balance the needs between data availability and security. Warren Gordon, business unit manager at Duxbury Networking, an Aruba technology distributor, said, "To survive and thrive in this digital age, enterprises have only one choice: modernize their network infrastructure. Taking the first step toward modernization is by exploring the seven key requirements for designing and deploying next-generation edge-ready networks."

[[409654]]

Requirement 1: Prepare for IoT and Wi-Fi 6, delivering data center-class performance at the edge

According to Ericsson, 3 billion new IoT devices will be deployed by 2023. Enterprise networks are being overwhelmed by IoT devices and bandwidth-hungry applications. Given the advent of Wi-Fi 6, access points can support more devices with up to four times the throughput, and traditional switches at the edge will inevitably be in short supply.

Modern networks require switches based on non-blocking architectures to deliver line-rate performance and provide a simple, powerful solution for high availability. These features, while common in data center infrastructure, become even more important at the edge where tolerance for performance issues is near zero.

“We need to optimize the use of every switch port to provide a seamless, best-in-class user experience, take full advantage of Wi-Fi 6 and IoT, and help prevent bottlenecks, ensuring there is always enough bandwidth for critical applications,” said Gordon.

Requirement 2: Strong investment protection to adapt to future needs

Gartner points out that by 2022, 95% of enterprises will still not know the connectivity requirements of all IoT devices accessing their applications. The average life cycle of a network switch is 5 to 7 years. Choosing an infrastructure that can provide strong investment protection is critical to supporting current needs and unpredictable future needs.

"While your edge network may only require 10GbE uplinks at the access and aggregation layers today, IoT and Wi-Fi 6 may force you to expand to 25GbE or even 50GbE uplinks. Gordon noted, "You need to invest in solutions that can provide flexible uplinks and support multi-gigabit Ethernet to help increase speeds on existing switch ports and existing cabling, keeping costs down and limiting disruption to your business."

Requirement 3: Use high-density, always-on POE to reduce costs

For any enterprise that is deploying wired IoT devices and Wi-Fi 6, having reliable, high-powered Ethernet (PoE) is a must-have.

High POE power on each switch port reduces costs by significantly reducing the additional cabling required to power these modern devices. Always-on POE ensures that critical devices, such as surveillance cameras or healthcare monitors, remain powered, even during scheduled reboots or firmware upgrades. In addition, using hot-swappable modular power supplies will help increase available POE power without expensive switch hardware replacements.

Requirement 4: Simplify network security with dynamic segmentation

80% of IT organizations report that they find IoT devices on their networks that are not installed, secured, or managed. The network edge must be intelligent enough to securely connect all types of devices and users. Relying on VLANs, ACLs, and subnets to properly segment the network is too time-consuming and error-prone. Instead, look for a network solution like Dynamic Segmentation that automatically enforces user- and device-based policies across wired and wireless infrastructure.

Centrally managing and enforcing network policies saves time, reduces errors, and strengthens security while ensuring that all types of users (guests, employees, or customers) have easy and reliable access to the right network resources.

Requirement 5: Simplify operations for network teams of any size through automation

According to Gartner research, automation can reduce human errors by more than 75% while increasing operational activities.

Many enterprises have adopted DevOps principles to build and release new or updated applications at a faster pace. To maximize the success of these efforts, the network needs to be as agile as the software it supports.

"We recommend that you build the right foundation for your modern network with an operating system that provides a cloud-like operating experience and enhances your automation framework," Gordon said.

For large enterprises, the key is to open APIs and build custom scripts to integrate with third-party systems. Teams with DevOps practices should consider extending existing automation platforms to support network-related workflows. Small teams that lack dedicated development resources should seek solutions that simplify common tasks, where compliance with network policies and device dependencies can be immediately verified.

Requirement 6: Use distributed analytics to quickly detect and resolve issues that affect users

39% of issues were discovered and reported by end users.

When network problems occur, operators need immediate insight into where and why the problem occurred. However, traditional network monitoring and troubleshooting methods are highly reactive and resource-intensive. As a result, problems that affect users often persist longer than they do.

Gaining real-time visibility into every network node is critical, especially in remote locations with little to no on-site technical support. Built-in intelligence can pre-process telemetry data of interest at the edge, tie it to possible root causes, and send alerts with all relevant diagnostics to centralized operations teams. This enables IT departments to detect, respond to, and even pre-emptively resolve issues that affect network health.

Requirement 7: Use a single operating model to drive operations from edge to data center

71% of IT budgets are spent on “running the business”.

Traditional networks are hampered by fragmented operations, resulting in insurmountable complexity, higher operating costs, and management headaches, including:

• Different architectures, operating models, and management tools across various network domains, often managed by different teams.
• Separate operating systems with different software requirements often require complex, subscription-based licensing plans.
• Implementing a single, end-to-end operating model can.
• Reduce complexity with a unified, end-to-end architecture spanning edge access to the data center.
• Reduce costs by replicating network designs in multiple areas.
• Facilitate closer collaboration between traditionally siloed network teams by leveraging a common operating system and toolset.

Digitalization of enterprise edge operations is the key to unlocking new growth opportunities. Enterprises can stay ahead of the competition by deploying an edge network that meets today’s technology demands while providing the scalability and flexibility to meet future needs.