3. Working with an ecosystem that is still in its infancy

[rakhirhif8963]

In his practice, which focuses on IoT security for medical devices, optimizing the costs and benefits requires a detailed look at workloads. “Sending all the network traffic collected on IoT devices to the cloud would require an unrealistically high amount of bandwidth — and would create data security issues. So the edge devices process it locally and send the resulting insights to the cloud,” Linden says. “But much of the computation, such as calculating risk scores for devices, can easily be done in a traditional cloud model.”

The expert advises to conduct an honest internal assessment of the costs and benefits of potential edge use cases — and then focus on those where performance/latency (or other goals) are truly critical. As with the cloud, an all-or-nothing approach should not be used.

François Baldassari, CEO of Memfault, cites a set of fundamental problems that exist, from a software perspective, everywhere - at the edge, in the cloud, or on traditional endpoints:

How do we deploy software and track versions?
How do we monitor performance and what is ghana mobile database strategy?
How do we track defects?
How do we detect and mitigate security risks?
There is already a proven set of answers to these questions for cloud and cloud-native applications. Baldassari points to the rise of systems reliability engineering (SRE) as a field that essentially exists to address these and other issues in modern software and infrastructure. DevOps, DevSecOps, GitOps, and other disciplines also offer overlapping approaches to solving similar problems.

While these practices may extend to edge deployments, the edge ecosystem itself is still relatively nascent compared to the cloud. “The ecosystem is still immature, and there are a number of incomplete, heterogeneous solutions in the marketplace right now,” Baldassari says.

Like cloud computing before it, the edge computing ecosystem will evolve over time. In the meantime, IT leaders should be extra careful when choosing platforms, tooling, and other requirements. “They should work with their vendors to understand what their reliability strategy is, what SLAs they offer, and what tools they provide IT teams to monitor edge devices on the corporate network,” says Baldassari. “If nothing else, they should use appropriate network policies to tightly control and limit what edge devices can do on the network.”

4. Focus on security of edge infrastructure and applications