Pluribus Networks Announces Netvisor 2.2
Pluribus Networks announced today the next evolution of its network hypervisor, Netvisor 2.2. The new release enhances enterprises’ ability to support complex application environments via an open architecture that leverages Netvisor’s bare-metal vFlow programmability and multi-threaded APIs. This results in ASIC-level network performance with full application-level analytics without the need for expensive, complex redesigns of ‘brownfield’ deployments.
Along with this new release, the company is announcing that Lucera, one of its customers running Netvisor in its on-demand, high-performance infrastructures, released a new Securities Technologies Analysis Center (STAC) benchmark scores showing the McObject eXtreme DB setting new scores in five of 17 STAC M3 benchmarks in a four- node distributed calculation.
“The STAC test is very important to us and our customer base as it continues to prove that you can have high performance compute on-demand. The Pluribus F64 is the backbone of Lucera and its high-performance converged switching technology helped us achieve these results,” said Jacob Loveless, CEO of Lucera. “When we decided to use the Pluribus Netvisor hypervisor, we looked at the opportunities that arise from tightly coupling the network and the compute. This new paradigm in backbone switching and the increased performance that came with the ability to program the network like a server have given our engineering team unprecedented flexibility to design high-performance solutions for customers.”
“The new Netvisor release, supporting both Intel and Broadcom switching, allows new cloud deployments to take advantage of Netvisor’s distributed programming environment,” said Sunay Tripathi, Pluribus co-founder and CTO. “Netvisor fuses the physical and virtual network using Switch Hardware Offload, critical in realizing the software-defined data center.”
New Netvisor capabilities include enhanced vflow programmability with ‘C’ and Java APIs, Active-Active Virtual Link Aggregation Groups, Rapid Spanning Tree Protocol, and the Virtual Router Redundancy Protocol. Manageability enhancements include vPort, the ability to track port attributes across the fabric over a specified time range, as well as the ability to visualize fabric services via the Pluribus vManage GUI. OpenStack extensions include integration and automation features designed to drastically reduce the time to power up racks of VM and storage, as well as Horizon-based NPU-based QoS SLAs for tenant networks and vFlow functionality, permitting customers to program and control application flows throughout a cluster of switches in a fabric.
Pluribus also announces its new high-capacity 40 GbE Network Computing Appliance, the E28Q, powered by Netvisor 2.2. The new appliance, based entirely on merchant silicon switch technologies, extends the Pluribus Netvisor into high-capacity 40 GbE configuration spines that are increasingly being deployed with fixed rather than modular switches.
“We are currently trialing the Pluribus Network Computing Appliances with Netvisor, and expect that the features announced today will help facilitate our unique service deployment,” said Sanjay Uppal, CEO and co-founder of VeloCloud.
