Pushing IoT to the Edge

Edge computing, a network system that allows data produced by internet of things (IoT) devices to be processed closer to where it is created instead of sending it across long routes to data centers or clouds, is gradually becoming the norm amid the rising adoption of IoT.

In an email interview with Networks Asia, Vertiv’s Tony Gaunt, Senior Director Colocation, Cloud & BFSI Asia, noted that the main benefits of edge computing lie in its ability to minimise latency as the hardware is deployed where data is collected. “As such, processing data is much faster as compared to the cloud. Industries that require real-time processing and time-sensitive decisions are expected to be early adopters of edge computing.”

In the interview, Gaunt answers questions about the impact of edge computing on security, analtyics, and current infrastructure. Gaunt also talks about the challenges that are currently holding organizations back from embracing edge computing.

Excerpts of the interview follow:

As we move to the Cloud and now the edge, does the burden of security move as well? How far or porous has the perimeter become? Is edge computing really the future of the internet?

The burden of security doesn’t move.  There will be an equal need to protect the many devices that will lie on the edge of the network as well as the backend infrastructure. Today, IoT devices that are already on the edge may not be capable of running security solutions because they are designed to be lightweight and affordable so there isn’t enough computing power to be secure. However, this approach also means that such devices remain the vulnerable point for any IoT implementation.

With edge computing, we will see such devices re-designed with security as top of mind and we will be able to do more with IoT now that security is a concern that is being addressed.

It should be made clear that edge computing will not displace cloud computing completely. They are two very different resources that complement each other. For instance, you would still require the cloud to push certain settings or big data findings into the edge devices. However, data that requires a huge amount of storage capacity or that is not time-sensitive would likely be redundant for edge devices. As such, this sort of data would be better off stored in the cloud.

At present, our current cloud infrastructure will not be able to handle the massive amount of data with the explosion of IoT devices. To support such demand, we will either see major developments for the cloud infrastructure or the implementation of new infrastructure such as edge computing. The future of internet is not cloud vs edge computing but rather how different approach complements each other.

How hardened are current devices and endpoints at the edge to deal with modern threats? Do we still need to run containers to secure enterprise apps and data?

The trouble with majority of IoT devices is that they may not have been designed with a security-first approach. This is because manufacturers have other priorities in mind such as the pressure to release a product first. Such an issue isn’t edge specific and applies to cloud only infrastructure too. Furthermore, such devices are designed to trust other devices without verifying their credibility. That said, it could be connecting to a device that’s malicious.

The main purpose of containers would be to help software run reliably when moved from one computing environment to another such as from the cloud to the edge or vice versa. During the migration process, problems may arise because of the change in software environment or the difference in security policies or network topography. With containers, organisations can ensure that no security loopholes arise during migration. At the same time, enterprises will be able to design and apply security policies that trigger alerts and execute preventive measures and responsive measures when a threat is detected in a container.

Do infrastructure requirements change as we move towards the edge? How are we moving from on-premise to the Cloud and then the Edge?

Just as how organisations required a change of infrastructure when moving to the cloud, moving to the edge is the same too. However unlike the cloud where we switched to a then, unfamiliar centralised network model, moving to the edge will not be on unfamiliar grounds as we will essentially be heading back to the decentralised network models seen during the late 1990s and early 2000s.

The key challenge for organisations would be to find the necessary system compatibility that would support the hybrid model of cloud and edge and whether they have the right people who are familiar with operating such a model.