Aruba boosts AI capabilities to help enterprises manage, troubleshoot the edge

Aruba is adding AIOps features to its Edge Services Platform (ESP) to help customers automate everyday tasks, shrink the time needed to find and fix problems, and increase edge security controls.

Rolled out in 2020, Aruba ESP analyzes telemetry data generated from Aruba Wi-Fi or network switching gear and uses it to automatically optimize connectivity, discover network problems, and secure the overall edge environment. ESP builds a data lake of a customer’s data center, campus, and SD-WAN switch information, and it combines that data with statistics from billions of data points generated daily by Aruba devices worldwide.

ESP can then apply its AI and machine-learning algorithms to troubleshoot issues before they become problems, said Larry Lunetta, vice president of marketing for security solutions at Aruba, Hewlett Packard Enterprise’s network subsidiary. ESP is also part of Aruba’s core network infrastructure management console, Aruba Central.

With this upgrade to ESP, Aruba is adding more AI-based features, which are available now, that are designed to help IT teams more easily manage some of the mundane tasks they face every day.

The newly added Aruba Client Insights feature, for example, will let customers discover what devices – switches, servers or endpoints – are on their enterprise network. It can then automate access privileges and monitor device behavior. Customers can see traffic flows and troubleshoot problems more quickly, Lunetta said.

“Client Insights is especially important for organizations that have a lot of attached IoT devices at the edge they may or may not even know about,” Lunetta said. “We discover components based on telemetry that we’ve built into the infrastructure – there is no separate collector, no separate agent required. Device information comes natively into Aruba Central and ESP, and then we build profiles, based on static and active attributes like duty cycles, amount of traffic produced, and details about what else the device talked to.”

Customers then can tag, define access control, and securely assign a role to a device, which lets IT feel much more comfortable controlling that device, Lunetta said.

Another new feature, Firmware Recommender, can determine whether an organization’s wireless network is running the most current versions of software and can recommend software fixes and upgrade solutions quickly.

Customers have multiple generations of equipment, Lunetta said, and it’s difficult to keep up with everything. Firmware Recommender removes that complexity. It takes the guesswork out of upgrading or changing software on devices, and it helps ensure new features and fixes are implemented quickly.

Another new feature that reduces the burden on network teams is Automated Infrastructure Predictions. It uses Aruba’s AI Assist feature to identify potential hardware and software failures and recommend or automate firmware upgrades or hardware replacement before a problem occurs.

Lastly, Aruba added Spanish language support. The same built-in natural language search function in Aruba Central now supports queries and responses in Spanish, Lunetta said.

While adoption of AIOps technology is in the early stages, experts say it will play a key role in the way complex enterprise environments will be managed in the next few years, and many Aruba competitors, such as Cisco, Extreme and Juniper, are also working on AIOps offerings. (Read more: Edge computing moves toward full autonomy)

Organizations need to leverage AIOps to effectively manage highly distributed and complex network environments, wrote Bob Laliberte, principal analyst at Enterprise Strategy Group, in a new research note about AIOps.

AIOps offers streamlined troubleshooting; in an always-on world, it is imperative to fix any problem that arises as quickly as possible and not waste precious time trying to find the root cause, according to Laliberte. 

“AIOps identifies root causes, provides recommendations, and even automates the remediation, depending on the problem or severity,” Laliberte stated. “Ideally, this would eventually lead to an environment that is proactive, continuously self-analyzes, and provides optimization insights when required, focusing on preempting problems, not just finding them. The goal is to reduce the volume of trouble tickets and time spent answering them.”