Extending the fabric to the campus: a software-defined head-office network for one of Ethiopia’s largest banks
The bank’s head-office campus rebuilt on the same software-defined principles as its datacenter — high-density core and access switching with EVPN/VXLAN, cloud-managed automation, and three years of support — awarded to eSafe through open competition.
The challenge
Three years after eSafe delivered the bank’s Tier III-ready datacenter, the contrast between the datacenter and the rest of the estate had become the problem. The datacenter ran on a modern software-defined fabric — automated, telemetry-rich, consistent. The head-office campus — the network that connected thousands of staff, meeting rooms, branches’ head-office functions, and the bank’s operational departments to that datacenter — still ran on a traditional switched architecture: layer-2 domains stitched together with spanning tree, per-switch configuration, and a management model that made every change a manual, error-prone task.
The consequences were practical. Campus outages could isolate entire floors. Moves and changes took days. Segmenting a department for security or compliance meant re-engineering VLANs by hand. And the operations team was effectively running two networks with two different skill sets and two different toolsets.
The bank’s target was clear: bring the campus onto the same software-defined foundation as the datacenter — one architecture, one automation platform, one operating model — with the port density, throughput, and resilience a large bank’s headquarters demands, and with multi-year support so the platform would be maintained, not just installed.
Our approach
- 01
Leverage what already worked
Because eSafe had designed the bank’s datacenter fabric, the campus design started from a known-good foundation: the same overlay principles, the same management platform, the same operational conventions. This is the compounding value of an integrator that has already earned the datacenter.
- 02
Design for density and resilience
The campus was engineered as a two-tier design — high-capacity core switching with dual redundancy, and high-density access switching to the floors — sized from the bank’s actual port counts and traffic profiles, with capacity headroom for growth.
- 03
Automate from the first switch
Every access switch was provisioned from templates through the central management platform, so configuration was consistent by construction. The bank’s operations team was involved from the design phase, because the goal was a network they would run.
- 04
Phased cutover by floor and function
Migration proceeded floor by floor, function by function, in agreed windows, with parallel connectivity maintained until each segment was verified. Head-office banking operations were never interrupted.
- 05
Deliver support as part of the design
Three years of hardware and software support, with defined next-business-day replacement, was built into the engagement — the platform arrived with its lifecycle already planned.
The solution
Software-defined campus architecture
- High-performance core switches with high-speed uplinks and full redundancy — the campus backbone, engineered so that a core failure is a non-event.
- High-density access switching across the head-office floors, delivering wired connectivity at scale to staff and endpoints.
- EVPN/VXLAN across the campus — the same overlay technology used in the datacenter, giving the bank flexible segmentation and consistent policy across campus and datacenter, without stretched layer-2 fragility.
- Fabric-grade optics and cabling engineered for the throughput and distances of a large headquarters.
One operating model
- Cloud-based network management for the entire campus estate — intended-state configuration, template-driven provisioning, change control, and lifecycle management from a single platform.
- Unified with the datacenter — campus and datacenter now share the same automation approach, the same telemetry, and the same skill set, so the bank’s engineers operate one network, not two.
- Streaming telemetry and observability, so campus health and drift are visible continuously.
Delivery and lifecycle
- Implementation services — design, installation, configuration, migration, and commissioning by eSafe’s certified engineering team.
- Three years of hardware and software support with next-business-day hardware replacement — the bank’s campus is covered for the platform’s operational life, not just its launch.
The result
The campus network was delivered on schedule and brought the bank’s head office onto a modern, automated, software-defined foundation — unified with the datacenter eSafe had built three years earlier.
The award itself is a result worth stating: this was an open, competitive procurement, and the bank’s evaluation found eSafe’s proposal the strongest. For a bank that already had years of experience with eSafe’s delivery, choosing eSafe again for its next major network was a decision made with full information.
Together, the datacenter and campus projects mean the bank’s core infrastructure — from the server rack to the desk — now runs on one architecture, delivered and supported by one partner.
Why it worked
One overlay technology, one management platform, one operating model across datacenter and campus — fewer skills to maintain, fewer ways to fail.
Automated provisioning made hundreds of ports consistent by construction.
Three years of support was part of the architecture, not a line item added later.
Client identity withheld under our confidentiality commitments. Details, quotations, and outcomes are drawn from the client’s formal project acceptance and recommendation letters, held on file.