With networks rapidly expanding through edge device proliferation, managing these devices is becoming more arduous. I had the opportunity to sit down with Sudhir Reddy, CTO of Esper, to discuss the growth in edge device deployments and how bringing DevOps, a process discipline normally associated with software development, could help solve a number of today’s edge challenges.
Esper concentrates on device management, in this case, enterprise edge devices distributed across a range of business locations. Great examples include airport kiosks displaying advertising or passenger information, restaurant POS systems or back-of-house displays, or even medical input systems targeted at patients and providers.
Previously, device management was a center-out motion with centralized servers and unsophisticated endpoints. But this model has given way to more engaging, sophisticated endpoints that must be responsive and interactive without relying on centralized updates. When staring down 50,000 or more endpoints in a network, central management is no longer tenable. Managing a dozen devices in a single restaurant might fit a centralized model, but a franchised model with thousands of worldwide locations in different time zones simply will not scale without some level of edge autonomy.
In this model, application and data become critical as organizations need to be able to view details from a high level across the whole system and drill down to the most granular details of a single device. Security, which is always at the top of mind nowadays, is essential when dealing with edge devices. Companies need to lock down edge devices in every way imaginable because most of these devices feature interfaces like USB that provide an entry point for attackers.
Many people associate DevOps with software, but the concepts espoused by DevOps are part of many disciplines beyond software development; DevOps is about workflows, processes, automation and operationalizing change management and control. It’s more of a systemic framework than a coding paradigm, which is why areas outside of software development, like cloud infrastructure management, have raced to embrace DevOps principles. However, the device world has been slow to embrace DevOps, creating opportunities for companies like Esper. With DevOps in device management, tools like Esper can define the endpoint devices from a managed configuration profile and then deploy this desired state to all the endpoints worldwide. Most importantly, it can then maintain that state, enforcing security and preventing endpoint modification of either devices or applications.
With regulatory and compliance considerations top of mind in the face of cyber threats, businesses need the ability to not only protect these edge devices, but also report on them to either auditors or regulators. DevOps really simplifies this process.
DevOps at the device level enables you to operate at scale, building pipeline, processes and flows that enable applying policies, while still controlling deployment down to the device level. Since organizations already use DevOps with software, extending that process model to hardware makes perfect sense.
Looking forward, we’re seeing more artificial intelligence (AI) being deployed at the edge, where it can be more responsive to users. So, as organizations extend their DevOps to devices with a tool like Esper, they can also think about how this model will be extensible for them with AI.
With the increasing importance of edge devices and the complexity of managing this expanding universe, it only makes sense that we turn to proven methodologies like DevOps to help keep edge devices running smoothly, updated and compliant, just as we have for software.
