Pluribus Networks is now Arista Unified Cloud Fabric

Unified Cloud Fabric for Multi-Site Data Center Fabrics and DCI

Unifying and simplifying networks across distributed data centers and clouds

From Data Center Interconnect to Unified Multi-site Fabrics

The Pluribus Unified Cloud Fabric provides flexible solutions to match your multi-site data center requirements.

For data center operators needing to connect existing data centers with a traditional data center interconnect (DCI) approach, a Pluribus fabric can provide seamless, transparent interconnection of two or more data centers over any existing DCI transport or IP network infrastructure. In this scenario, the DCI fabric stops at the edge of the connected data centers, which may still operate as islands that are not fully unified.


For operators who want to move beyond simple DCI, the Unified Cloud Fabric provides a powerful solution to truly unify distributed data centers and allow them to be operated as simply as a single data center, enabling greater agility, application availability and mobility, and capex and opex efficiency.

How it works

The Unified Cloud Fabric is a distributed application powered by the Netvisor ONE operating system running on top-of-rack leaf switches and in server-based data processing units (DPUs) in distributed data centers. These diverse network edge platforms are connected into a single unified fabric that stretches across any geographic distance and any existing wide area network. The fabric incorporates a distributed, controllerless SDN control plane, providing end-to-end network automation, distributed virtual network services with secure multi-tenant segmentation, and built-in monitoring and analytics for pervasive network visibility.

Features and Benefits

Comprehensive automation for simplified operations

Unlike any other solution, the Unified Cloud Fabric automates both the underlay network and overlay network virtualization with built-in monitoring and analytics. This enables dramatically simplified operations, from streamlined day 0 deployment and one-touch automated fabric-wide service and policy provisioning to end-to-end fabric-wide monitoring and visibility. The result is increased agility, lower operational costs, and more time for operations staff to focus on higher priorities.

Cost-, space- and power-efficiency

The controllerless SDN architecture running directly on open network switches and DPUs eliminates the need for multiple redundant controllers at every site, enabling dramatically better cost-, space- and power-efficiency – especially important for smaller data center sites.

Distributed security and segmentation

Deep network segmentation across the data, control, and management planes supports secure multi-tenant data center services and isolates applications and untrusted devices, including Internet of Things (IoT) devices. Microsegmentation and distributed firewall capabilities implemented in server-based DPUs enable a true zero-trust security architecture.

Pervasive visibility and insight

Built-in wire-speed monitoring, which requires no external hardware and has no performance impact, provides full visibility into physical and virtual networks – every port, every flow, every application, from start to finish.


A multi-site data center fabric creates opportunities to improve application availability and mobility while increasing opex and capex efficiency.

A Pluribus multi-site data center fabric enables geographically separated data centers to easily participate in any form of business continuity and disaster recovery (BCDR) strategy. Unlike traditional data center interconnect approaches, the Pluribus multi-site data center fabric allows all resources to appear as if they exist within the same “local” infrastructure. When the physical delineation between data centers is removed, critical business information can be backed up, replicated, or accessed in real-time, and active-active or active-standby strategies can be implemented much more simply. In the event of failures or IT maintenance schedules, workloads can be easily moved or recovered in any location.
A Pluribus multi-site data center network fabric enables dynamic and virtualized processing to span across all available resources, regardless of location. This increases the utilization of each resource and provides a mechanism to monitor the utilization of local versus remote resources, providing an opportunity to add capacity in the most-needed locations.
A Pluribus multi-site data center fabric completely eliminates the boundaries between data centers and the resources that they house. Workloads can move among sites in the fabric for capacity or availability reasons with no IP readdressing needed, and continue to be accessed from other sites as if they were local.

Start unifying your multi-site data center today.