Pluribus Networks today introduced the seventh generation of its networking platform that supports 400 Gb/s switches and enables visibility in network flows — including those inside Kubernetes clusters.

“The main thing we're trying to do is make the life of the network operations team easier … to give them the tools they need to truly understand what's going on in the network,” Pluribus CMO Mike Capuano told SDxCentral. 

Pluribus’ Adaptive Cloud Fabric federates multiple geographically distributed switches so that they appear as one logical switch. When paired with its Netvisor One operating system, the platform deploys a layer 3 service across a 256-switch data center fabric with one or two commands, a feat that requires more than 5,000 commands in traditional networking architecture, the company claims.

With the release of its revamped Netvisor One and Adaptive Cloud Fabric platform, Pluribus rolled out three new fabric capabilities: KubeTracker, FlowTracker, and Virtualized Packet Broker.

The KubeTracker is designed to “empower network operations teams with Kubernetes visibility,” Capuano explained. “The challenge that network operations team have right now is they don't have visibility into where these containers are, which container’s talking to which other container, what application each container is associated with.”

Because of this, Pluribus introduced KubeTracker to provide visibility into east-west traffic between containers and enable NetOps teams to correlate containers with applications.

“What we do is the fabric actually subscribes to the APIs of the Kubernetes server, and we ingest all that information” and use it to offer visibility inside Kubernetes clusters, Capuano said.

On top of Kubernetes, Pluribus also intends to “track every flow across the fabric without needing any external equipment” with its FlowTracker capability, according to Capuano.

The FlowTracker offers application-level network flow telemetry based on Broadcom’s Trident3 and Trident4 programmable Ethernet switch chips that are built into data center switches, Pluribus claims.

In addition to bringing more visibility into cloud-native applications, Pluribus also aims to reduce costs for customers.

It does this via Flowtracker, which Capuano said eliminates the need for and costs associated with separate taps and tap aggregation capabilities.

Plus, Pluribus' virtualized packet broker service cuts the costs associated with traditional network packet broker tools. The software-defined service runs on the same network fabric that carries production cloud data center traffic as the packet brokers, Capuano explained.

Additionally, the virtualized packet broker can span multiple geologically distributed data centers including the edge ones that a packet broker typically cannot accommodate.

Along with the platform upgrade, Pluribus also announced support for Dell Technologies’ 400 Gb/s white box switch.

“We timed our investment in bringing to market the most advanced 400 Gb/s platforms at the right time, which is this Dell PowerSwitch Z9432F-ON,” Capuano boasted.

Dell'Oro Group analyst Sameh Boujelbene concurred on the timing. “We see the broader market start to adopt 400 Gb/s at the end of 2022 with a more meaningful volume starting in 2023,” she said in a statement.

The support for higher-performance switches will allow customers “to scale the data center in terms of higher speed capacity for richer applications and better application performance,” Capuano said.