When Henry Ford looked to apply novel motor technology to our ever-growing need for transportation, it looked risky.
A bank advised his attorney not to invest: “The horse is here to stay but the automobile is only a novelty – a passing fad.”
We know how that turned out.
The era of faster horses
The architecture of most campus networks was defined in the client-server era, circa 2000, when the primary need was to increase wired access speeds and internetwork more locations. Networks were managed via command line interface (CLI), and other than Power-over-Ethernet, the main improvements were step-ups in access speed: 10Mbps to Fast Ethernet to Gigabit. This era has lasted for two decades.
The era of mobile-cloud and IoT
Today’s problems are different. Mobility has changed the nature of access, requiring support for multiple devices per person with an unpredictable location, new traffic patterns, unknown ‘things’ on the network, and a new threat landscape without a trusted perimeter.
Cloud has also changed the nature of applications towards centralised architectures, shifting user application traffic from client-server to app-cloud.
The Internet of Things (IoT) is everywhere – beyond mobile personal devices to ‘things’ like printers and video cameras, plus a new wave of connected devices from HVAC to sprinkler systems to physical device tracking and more.
The campus network is forced to take on three roles: (1) access, (2) IoT gateway, and (3) wireless backhaul. In the core of the network, in particular, the old approach has become brittle. Operations relying on CLI for manual configuration and troubleshooting are error prone. Network visibility via log data requires off-line (delayed and incomplete) analysis. Inflexible networks require hard-coding to integrate with services.
An evident customer pain
The common challenge amongst enterprises today is that they are constrained by the decades-old core infrastructure that fails to provide the agility and flexibility they need. These pain points can be alleviated with deeper network visibility, intelligent troubleshooting, and automation of network tasks.
Three reasons we need automobiles, not faster horses
1. Visibility – what is happening?
Network visibility is more challenging than ever. Increasingly organisations are turning to expensive overlay networks to mirror traffic, and they’re still struggling to keep up with what’s happening. Inevitably this approach fails to provide real-time comprehensive visibility. IT managers need to know: “What is happening on my network right now?”
2. Troubleshooting – where and why is it happening?
Once the network has deeper visibility – articulated where and how the network operator needs – the next step is to aid the troubleshooting process itself. The network needs to do more to connect what is happening to where and why it is happening. It should provide intelligent insight to questions such as: “If our video conferencing service has poor quality, what else is going wrong at the same time?” and “When was the last configuration change?”
3. Automation – fix it fast!
Automation of network tasks is one of the most common objectives for the network team, but we’re still not there. New apps can be spun up in containers in virtual clouds, but campus networks are still manually stitching together SSIDs, VLANs, VRFs, QoS policies, and more. Networking experts spend a lot of time typing CLI and not enough time on the tough problems of business efficiency, IoT enablement, and more.
Automation requires full programmability, easy integration with industry standard tools, and a readiness for the modern world of micro-services and orchestration.
Put the horses out to pasture
Across the tech landscape, organisations need automobiles, not faster horses. To get there, the core of the network would need to change – offering native visibility, troubleshooting, and automation to help customers do more with less and deliver what the business needs with maximum efficiency.
Not everyone will be ready for a change. But change is coming!