Hewlett Packard Enterprise (HPE) today announced it is bringing additional artificial intelligence (AI) capabilities to its networking portfolio, following its acquisition of OpsRamp last year.
OpsRamp had previously been applying machine learning algorithms to optimize the management of IT operations, otherwise known as AIOps.
Alan Ni, senior director for edge marketing at HPE, said the core AI technologies developed by OpsRamp via a public preview are now being applied to extend the depth of observability that can be applied to the management of its Aruba networking portfolio, including wireless access points, switches, firewalls and routers.
In addition, HPE is surfacing additional AI insights along with providing access to any network device configuration engine that can now be applied to all devices connected to the HPE Aruba Networking Central platform along with 90 application programming interfaces (APIs) that can be used to extend the reach of that engine.
Finally, the digital experience monitoring (DEM) capabilities enabled by HPE are also being extended by natively integrating HPE Aruba Networking User Experience Insight (UXI) into the core HPE Aruba Networking Central platform. HPE Aruba Networking UXI sensors continuously monitor adherence with service level agreements (SLAs).
HPE has steadily been adding a range of AI models to its portfolio over the last six months. For example, a set of classification AI models have been trained using telemetry data collected from more than 4.6 million network-managed devices and more than 1.6 billion unique customer endpoints.
Earlier this year HPE also expanded its AIOps network management capabilities by integrating multiple large language models (LLMs) within HPE Aruba Networking Central along with additional security observability and monitoring capabilities enabled by machine learning algorithms.
As applications become more distributed, the need for deeper levels of observability into the networking services upon which that software depends has become more crucial. That requirement, however, requires greater dependency on AI models to surface anomalies indicative of performance issues or cybersecurity events that should be further investigated, said Ni.
The degree to which IT operations teams are relying on AI to optimize the management of IT infrastructure will naturally vary. However, given the inherent complexity of the networks being employed today, it’s apparent that most IT teams are no longer able to rely on manual processes to manage networks at scale. Those AI models are not likely to eliminate the need for IT professionals as much as they will reduce the overall toil encountered when, for example, trying to configure a fleet of devices.
Those capabilities should also serve to make networking operations teams more agile in an era where configuration change requests are made more frequently, noted Ni.
Hopefully, there will come a day soon when network management is automated to the same degree as the management of servers and storage. Unfortunately, it’s still not uncommon for changes to networking environments being made in days, when changes to the rest of the IT environment can now be consistently made in a matter of minutes.
