Pluribus NOS upgrades target Kubernetes, cloud fabric management

Pluribus Networks has significantly upgraded its switch-fabric software to provide a better handle on distributed, containerized enterprise-cloud resources.

The upgrades add three new fabric-monitoring capabilities—FlowTracker, KubeTracker, and Virtualized Packet Broker Service—to Netvisor One, the company’s virtualized Linux-based NOS that provides Layer 2 and Layer 3 networking and distributed fabric intelligence.

The NOS virtualizes switch hardware and implements the company’s Adaptive Cloud Fabric software-defined networking package. Adaptive Cloud Fabric operates without a controller and can be deployed across a data center or targeted to specific racks, pods, server farms, or hyperconverged infrastructures, the company said.

The first new Netvisor One feature, FlowTracker, adds wire-speed network-flow telemetry for every flow-type that can traverse Adaptive Cloud Fabric including TCP, UDP, ICMP, as well as infrastructure services such as DNS, DHCP, and NTP. Until now, Netvisor ONE and Adaptive Cloud Fabric could capture flow telemetry for TCP flows only.  

The software gathers network telemetry data from Ethernet switches built on Broadcom’s Trident3 and Trident4 programmable Ethernet-switch chips. Core networking vendors including Cisco, Juniper, Arista, Dell, HPE, and others support the chips.

10 competitors Cisco just can’t kill off

Integrating telemetry support eliminates the need for external test access points (TAP) and TAP-aggregation overlay infrastructure, wrote Alessandro Barbieri vice president of product management with Pluribus in a blog.  Network TAPs are hardware devices placed on network segments to monitor network traffic.

The cost of  TAPS can be daunting and often results in cost—benefit tradeoffs, where often TAPS are only installed at certain points in the network. FlowTracker eliminates that expense, captures flow metadata at every switch, and exports it to tools such as Pluribus’ UNUM Insight Analysis for spotting and fixing problems more quickly, Barbieri stated.

The new Pluribus software also adds KubeTracker, which monitors and manages visibility into traffic flows between containers. It correlates containers with applications, determines on which hosts they reside, and how they are connected to the fabric. 

The idea is to quickly identify and troubleshoot application availability issues, Barbieri stated. Until now, NetOps teams have not had visibility into these flows, making it difficult to verify application availability and network performance for containerized workloads.

“KubeTracker is a service that is supported on any switch that is part of a Pluribus fabric (enabled on any of the fabric nodes) to subscribe to the events published by the Kubernetes API server and make the entire fabric aware of how Kubernetes micro-services are distributed across the network and how their deployments dynamically change over time,” Barbieri stated.

The third new Netvisor One component is Virtualized Packet Broker that aggregates, filters, and replicates any flow from anywhere in the fabric and direct that data to one or multiple monitoring and security tools.

“What I find interesting is that Pluribus can deliver the Virtualized Packet Broker as a service of the Adaptive Cloud Fabric,” wrote Ron Westfall, senior analyst and research director with Futurum in a blog about the Pluribus news.

“This is key, as packet brokers are typically deployed as part of a parallel monitoring network consisting of lots of hardware including TAPS, which literally tap into the network to create mirrored data flows and TAP Aggregators which hand off aggregated traffic streams from these TAPS to dedicated packet broker appliances. This infrastructure can be incredibly costly to purchase, deploy, and maintain. Pluribus is looking to deliver this virtualized network packet broker as a virtual service running on the same fabric that is carrying production layer 2 and layer 3 network traffic. I see this solution as clearly differentiated from alternatives, particularly legacy TAPS implementations.”

In addition to the services upgrades, Pluribus said Netvisor can now run on Dell’s 100/400Gbe PowerSwitch Z9432F-ON, which is aimed at high-end data-center network-aggregation duties. Dell and Pluribus have worked together for a number of years to integrate data-center networking hardware and software that competes with offerings from HPE, Cisco, and others. 

Pluribus Netvisor ONE OS for the Dell PowerSwitch can be ordered directly through Dell.