At a glance
- Client
- A financial institutionName withheld for confidentiality
- Industry
- High Tech
- Scope
- Data centre network design, migration and automation
The challenge
The institution's three-tier data centre network was built for a few large servers, not hundreds of containers talking to each other. Bottlenecks appeared as workloads grew, and every change was a manual, after-hours task.
Development teams waited days for network changes.
Our approach
Design spine-leaf
A spine-leaf fabric gives every server the same short, predictable path to every other server.
Upgrade the backbone
High-speed links were sized for current and future east-west traffic.
Automate changes
Network configuration is defined as code and deployed through tested pipelines.
Migrate safely
Workloads moved rack by rack with rollback plans and no business downtime.
The solution
The data centre now runs on a high-capacity fabric suited to containerised applications. Network changes that took days are made in minutes through automation, with fewer errors.
Monitoring shows traffic flows across the fabric in real time.
Results
- Predictable performance for east-west traffic
- Network changes delivered as code
- Migration without business downtime
- Real-time visibility of traffic flows
Technology used
- Spine-leaf fabric
- EVPN-VXLAN
- Ansible
- Network CI/CD
- Telemetry
Facing a similar challenge?
Tell us what you are working on. We will walk you through how we would approach it, with no obligation.