Juniper brings AI to data center operations

Juniper Networks is bringing its AI-based management technology to its intent-based data center software, Apstra, in an effort to simplify design, deployment and operations. The move brings Juniper’s wired, wireless data center, campus and branch networking products under one common offering, which the networking vendor calls its AI-Native Networking Platform.

Central to the AI-Native Networking Platform are the firm’s cloud-based, natural language Mist AI and Marvis virtual network assistant (VNA) technology. Juniper’s Mist AI engine analyzes data from myriad networked access points and devices so it can detect, offer actionable resolutions, and fix anomalies and problems. Marvis can detect and describe countless network problems, including persistently failing wired or wireless clients, bad cables, access-point coverage holes, problematic WAN links, and insufficient radio-frequency capacity. 

By bringing AIOps with Mist and Marvis to the data center, Juniper aims to drive automation and greatly simplify configuration and management, said Christian Gilby, product marketing director with Juniper. In the same way that Marvis reduces network trouble tickets in the campus and branch, Juniper expects it to meaningfully drive down issues, spot problems and reduce resolution time in the data center, he said.

Since it bought Apstra in 2021, Juniper has been bolstering the software with features such as automation, intelligent configuration capabilities, multivendor hardware and software support, and improved analytics, with the goal of making the system more attractive to a wider range of enterprise data-center organizations.

Apstra works by keeping a real-time repository of configuration, telemetry and validation information to ensure a network is doing what the organization wants it to do. Companies can use Apstra’s automation capabilities to deliver consistent network and security policies for workloads across physical and virtual infrastructures. In addition, Apstra performs regular network checks to safeguard configurations. It’s hardware agnostic, so it can be integrated to work with Juniper’s networking products as well as boxes from Cisco, Arista, Dell, Microsoft and Nvidia.

In addition to adding Apstra to Mist’s purview, the vendor has added a few new features to the package: Marvis Minis, Marvis VNA for Data Center, and the ability to manage AI-traffic.

Marvis Minis set up a digital twin of a customer’s network environment to simulate and test user connections, validate network configurations, and find/detect problems without users being present and without requiring any additional hardware, according to Juniper. “Minis simulate end user, device, and application traffic to learn the network configuration to proactively determine network issues,” Gilby said. “Data from Minis is continuously fed back into the Mist AI engine, providing an additional source of insight for the best AIOps responses.”

Marvis VNA for Data Center is a central dashboard for customers to see and manage campus, branch, and data center resources. Using Marvis’ natural language and integrated generative AI, the VNA looks at everything from cabling and configuration to monitoring network links and other operational issues, Gilby said.

The Marvis conversational interface lets IT teams pose direct queries and it highlights anomalies and recommended actions in data center switching devices, virtual infrastructure, physical and logical connectivity, and security, according to Juniper. The idea is to enable faster root cause identification and issue resolution, the company said.

In addition, Juniper said the Apstra software has been upgraded to better process AI/ML traffic over Ethernet, including congestion management, load balancing and flow control. InfiniBand was the first AI connectivity option, because many GPUs support it, but Juniper thinks, ultimately, Ethernet is the way AI networking development will go, Gilby said.

Juniper is part of the Ultra Ethernet Consortium, which was founded last year to develop physical, link, transport and software layer Ethernet advances – in particular to handle the performance, scale and bandwidth required to keep up with AI demands. The consortium includes AMD, Arista, Broadcom, Cisco, Eviden, HPE, Intel, Meta and Microsoft Dell, DriveNets, Fujitsu Limited, Huawei, Nokia and others.

Along with tweaking the Apstra software to better support Ethernet, Juniper added high-density 800GE routers and line cards to its PTX Series of boxes.

“The PTX Series routers for data centers have been extended with the new high-density 800GbE PTX10002-36QDD fixed switch and new 800GbE line cards for the PTX10000 chassis. Built on Express 5 custom silicon, the largest PTX10000 chassis now supports up to 576 x 800GbE ports for high radix spine and super spine architectures,” wrote Jeff Aaron, vice president of enterprise marketing for Juniper, in a blog about the new boxes. 

“In addition, a new high-density 800GbE QFX5240 fixed platform based on the latest Broadcom Tomahawk 5 ASIC offers silicon diversity for high-performance, scalable AI data centers with power efficiency,” Aaron wrote.

As for availability, the Marvis Minis will be ready this quarter, and the Marvis VNA for Data Center in Q2. The QFX5230 and PTX10002 are available, and the QFX will be available in the second quarter, Juniper stated.