SIXUNITED SDN Network Delivery Solution: Eliminates 90% of redundant operations and increases link capacity by approximately fivefold!
Release time:
2026-10-22
In August 2026, SIXUNITED successfully completed the delivery of a network project for the production line of a certain smart manufacturing factory. When terminals on this smart factory’s production line operate in a centralized manner, they require concurrent transmission of large-capacity system images, which can generate sustained concentrated traffic within a short period, placing high demands on core forwarding, access uplinks, and server-side links. As the number of terminals and network devices increases, the network must also possess greater bandwidth capacity and sufficient room for future expansion.
Centering on this business scenario, the project adopts an SDN‑based management architecture tailored for cloud‑native campus networks. Key deliverables focus on the wired production network, leveraging capabilities such as Zero Touch Provisioning (ZTP), batch configuration, graphical monitoring, and automated alerting and inspection to reduce IT operations overhead associated with manual per‑device configuration and fragmented queries, thereby enhancing the efficiency of large‑scale production‑line network deployments and day‑to‑day operations. Following delivery, redundant configuration rollouts have been cut by more than 90%, core link utilization has dropped from 88% to 35%, and all operational‑maintenance checks can now be viewed on a single dashboard.
The SDN network delivery solution covers production-line terminal access, core forwarding, and server-side links.
ZTP zero-configuration automated deployment reduces repeat deployments by 90%.
Traditionally, each time a switch is connected, it requires cabling, login, configuration entry, and version verification—repeated for every device. When the number of devices grows into the hundreds, a significant portion of on-site time is consumed by these repetitive tasks. ZTP (Zero Touch Provisioning) was specifically designed to eliminate such redundancy: once a device is powered on and connected, it automatically discovers itself, comes online, and retrieves its basic configuration—eliminating the need for manual, device-by-device login.
SIXUNITED places this capability at the forefront of the entire delivery process—before equipment arrives on site, organizational structures, premises, equipment inventories, and basic configurations have already been set up in the controller. Once initial connectivity is established on-site, the remaining devices automatically obtain their online information and connect to the controller, enabling automatic device onboarding, management, and synchronized configuration deployment. Engineers only need to verify the process with the first device; subsequent devices can be connected in bulk, with versions, statuses, and link relationships centrally reviewed. Based on the scale of Phase I equipment, this eliminates the repetitive manual onboarding work that would otherwise be required for each individual unit. Reduced by more than 90% Equipment installation, version preparation, link verification, and business validation remain under the engineers’ meticulous oversight; the only reductions are in the workload associated with redundant data entry and individual device commissioning, while delivery quality standards remain unchanged.
ZTP Automated Provisioning: Completes six steps—device onboarding, information acquisition, controller connection, configuration deployment, and status verification.
Centralized configuration baseline, with consistent deployment to similar devices.
Configuring each device individually poses another hidden risk: repeatedly entering the same information can inevitably lead to occasional omissions or parameter discrepancies, complicating subsequent troubleshooting. The SIXUNITED team establishes configuration baselines based on device roles and deploys them centrally through the controller, ensuring that devices of the same type share identical baseline configurations. For any newly added or replaced equipment, they simply reuse existing templates, eliminating the need for reconfiguration from scratch. Moreover, the controller can automatically detect device connectivity relationships, allowing engineers to verify device registration, configuration application, and operational status—all from a single interface—completing all checks in one go.
Visualized management—monitor the entire network’s status on a single screen.
Once devices are brought online, they gain a unified entry point for daily management. The device view provides a centralized display of online status, health metrics, and CPU and memory utilization, with detailed information on ports, optical modules, temperature, fan speeds, and power supply conditions. The topology view illustrates the relationships between devices and links, while the terminal view tracks online trends and traffic fluctuations. When an anomaly occurs, users can first assess the scope of impact from the overview, then drill down to individual devices for verification—eliminating the need to log in to each device one by one and perform manual troubleshooting.
Unified management of switches: health status, memory utilization, CPU utilization, and firmware version can all be viewed on a single screen.
Switch Health and Hardware Status
Wired Terminal Online Trend
Traffic monitoring identifies bottlenecks, and link capacity margins have expanded by approximately fivefold.
The controller continuously collects traffic data from each interface and displays historical trends showing inbound and outbound traffic variations over different time periods, enabling precise identification of capacity‑related issues down to specific links. Prior to the adjustments in this project, the Tx sampling rate on the 10‑Gbps uplink interfaces of the core switches was recorded at 1.10 GB/s, equivalent to approximately 8.80 Gbit/s—already accounting for 88% of the nominal bandwidth. The bottleneck was concentrated on the link segment connecting the access layer to the core aggregation layer.
Centralized Data Transmission Path
After the upstream link is upgraded to 25G, the nominal bandwidth increases to 2.5 times its original level, while the resource utilization for the same traffic volume drops to 35%. The available margin has expanded from approximately 12% to approximately 65%. This corresponds to an approximately fivefold increase in link margin. The controller continuously monitors the operational performance curves of four 25G uplink ports, none of which have approached their nominal capacity limits, thereby leaving ample headroom for future traffic growth. On the server side, 100G interfaces handle aggregated traffic, 25G links carry access‑to‑aggregation traffic, and 100G links serve as server egress points, with a clear two‑tier division of capacity. From bottleneck identification and capacity‑expansion implementation to ongoing monitoring, every step is supported by verifiable data in the controller, ensuring that all expansion decisions are firmly grounded in empirical evidence.
Original 10G uplink on the access side: Tx sampling rate of 1.10 GB/s
Observation of 25G Uplink Operation
Observation of 100G Server-Side Link Operation
Closed-loop alert inspection, with operations management at a glance on a single dashboard.
Where to view alerts and how to retain their results—these are all managed through a unified portal provided by the controller. Alerts are presented in a centralized manner, with some accompanied by potential root causes and recommended troubleshooting steps, helping engineers quickly narrow down the scope of investigation. Thresholds are maintained via the same unified interface, with optical power thresholds configured separately for each interface rate, ensuring that alerts are both accurate and actionable.
Intelligent Alert Analysis: Automatically identifies the root cause and provides step-by-step remediation recommendations.
Unified Alert Threshold Configuration
Centralized presentation of equipment inspection results
One-click and periodic inspections centrally monitor CPU, memory, temperature, fans, power supplies, and interface statuses; results are automatically archived and can be reviewed at any time, eliminating the need to log in to each device individually.
Delivery is not the end; the journey will continue.
This delivery has incorporated redundant go-live procedures, repeated data entry, individual device inquiries, and decentralized inspections into a standardized workflow, allowing the on-site team to devote more attention to the human‑intensive tasks of equipment installation, version preparation, link verification, and business validation. SIXUNITED has distilled this experience from the production‑line network delivery into a reusable methodology. How to bring devices online, how to monitor their status, and how to make scaling decisions—these can all be directly applied to similar scenarios. Delivery is just the starting point; this methodology will continue to deliver value throughout long‑term operations and maintenance.
recommend News