Cisco iconCiscoSep 2, 2026 ~6 min source read

Cisco Nexus One and VMware Cloud Foundation: a standards-based bridge between VCF and the data center fabric

Nexus One now interoperates with VMware Cloud Foundation 9.1 using EVPN-VXLAN and BGP to provide direct host-to-fabric routing, unify VM, container, and bare-metal domains, and remove edge choke points from private cloud deployments.

From fabric to cloud native: Cisco Nexus One for the modern private cloud

Share this story

Send the public story page.

Useful takeaways from this story.

VCF 9.1 peers directly with Nexus One EVPN Border Gateways via MP‑BGP EVPN and VXLAN, enabling dynamic route exchange instead of static VLAN/VRF stitching.

Nexus One uses standard EVPN/VXLAN protocols and runs on existing hardware, so the integration works off-the-shelf without bespoke fabric software.

A unified operating model gives network, cloud, and application teams consistent topology, tooling, and end-to-end visibility across heterogeneous private cloud domains.

# What changed

Cisco Nexus One now interoperates with VMware Cloud Foundation (VCF) 9.1 using open EVPN‑VXLAN standards. VCF's Networking Route Controller peers with a Nexus One EVPN Border Gateway using MP‑BGP EVPN, and ESX transport endpoints gain direct Layer‑3 reachability to the fabric via VXLAN. That replaces the older pattern of static VLAN/VRF mappings and centralized edge nodes.

# Why this matters

Private clouds commonly run virtual machines, Kubernetes, and bare‑metal servers together. Historically each domain used its own control plane and tooling, and connectivity between them relied on fixed edge devices and manual configuration. The Nexus One + VCF integration turns those static boundaries into dynamic route exchange. The result is fewer dedicated edge nodes in the packet path, simpler operations, and one consistent networking model across domains.

# How the integration works (technical summary)

  • Gateways using MP‑BGP EVPN.
  • Each VCF Transit Gateway maps to a Layer‑3 VNI in the fabric, preserving tenant isolation one‑to‑one.
  • EVPN Type‑5 routes exchange workload and external prefixes dynamically between VCF and the fabric.

This is built on standard, shipping Nexus One capabilities and requires no bespoke integration layer on the fabric side.

# Operational impact

Network teams keep a single architecture: the same topology, protocols, and configuration practices apply to virtual and physical workloads. Cloud and VCF administrators retain the VPC‑based self‑service model their users expect. Application teams can deploy VPCs and workloads without managing virtual appliances or manually mapping network constructs, because dynamic route exchange replaces the prior static coordination.

From a cost and performance perspective, host‑to‑fabric VXLAN routing maximizes throughput for VM and accelerator‑heavy traffic while eliminating edge node infrastructure that previously added capital and ops overhead.

# Standards and choice

The integration relies on EVPN and VXLAN standards already widely adopted in data centers. Cisco points to its long history of contribution to BGP EVPN and overlay standards and positions Nexus One as implementing those protocols on existing hardware. That gives customers the option to adopt the Nexus One operating model without vendor‑specific proprietary stitching.

# Practical considerations for adoption

  • Verify VCF 9.1 compatibility with your Nexus One release and existing hardware.
  • Plan mapping between VCF Transit Gateways and fabric Layer‑3 VNIs to preserve tenant isolation.
  • Confirm orchestration and visibility workflows so cloud and network teams share consistent operational views.
  • Assess edge node elimination for cost and scaling benefits, and identify any remaining centralized services that still require special handling.

# Bottom line

More context around this story.

AI-Ready Private Cloud with Cisco and VMware
Vmware iconVmwareSep 8, 2026

AI-Ready Private Cloud with Cisco and VMware

<p>An AI-ready private cloud is not simply a private cloud with GPUs added to it. What determines whether a private cloud platform can actually serve <strong>AI workloads</strong> effectively is everything built around them: how the fabric carries traffic, how the tenancy model lets teams consume capacity, and how poli

Loading more related stories...

Keep reading in the app

Open the app view to save this story, compare related coverage, and continue from the same source.

Open in app