The conductor and the network architect

For financial services institutions, the challenge is to innovate fast, but securely. They need the flexibility to launch new services – such as digital currency and AI investment advice – in a standardised and automated fashion.

This puts intense pressure on their networks. Whether they are connecting within datacentres, between datacentres, or between datacentres and the cloud, they need to be resilient and efficient.

Many institutions build heterogeneous networks, drawing on the best technologies available from a range of suppliers. It’s a good way to avoid vendor lock-in and reduce the risks of a single vendor failing.

But this comes with its own challenges. Different vendors use different protocols and manage their systems separately.

Working across multiple vendors can be costly, too. Figuring out who’s responsible when a fault occurs can be time consuming, and collaborating across multiple suppliers can be difficult.

Bring on the conductor

To help financial services institutions with their network challenges, Huawei has created the Xinghe Intelligent Ultra-Resilient Financial Data Centre Network Solution. It enables virtual network provisioning across multi-vendor hardware, while addressing the challenges of integration, security and management in heterogeneous networks.

“It functions like a conductor orchestrating instruments from different manufacturers into a harmonious symphony,” says Arthur Wang, President of Data Centre Network Domain, Huawei Data Communications Product Line. “Equipped with a digital map that visualises the entire data centre network (DCN), it delivers a unified operational view and control plane, enabling multi-vendor networks to operate as a single cohesive system.”

The solution is based on CloudEngine series switches and the iMaster NCE management, control, and analysis platform.

“The CloudEngine switches not only deliver high-performance connectivity with 400GE and 800GE capabilities but also feature rapid failover mechanisms to ensure business continuity even during partial network disruptions,” says Mr Wang. “Additionally, they are equipped with intelligent lossless network technology and multi-layered security functions.”

Mr Wang says that iMaster NCE serves as the neural centre of the network, delivering three mission-critical capabilities:

• Network-wide topology and policy visualisation enabling automated configuration.
• Automatic synchronization of security policies across multi-vendor environments.
• AI-driven self-healing with mean time-to-recovery under 5 minutes.

“Think of it as an intelligent orchestrator that not only understands each device's characteristics and requirements but also anticipates potential issues and initiates corrective actions proactively – much like a precision-engineered control system maintaining optimal network health,” Mr Wang explains.

Introducing the network architect

The solution provides insight and analysis into the device layer, network layer, protocol layer, overlay layer and service layer. By integrating more than 70 network performance indicators and more than 70 application performance indicators, the solution can help to identify issues in the application or network quickly. “Traditionally, network troubleshooting can take hours or even days. Our system can reduce this process to minutes, greatly reducing service downtime,” says Mr Wang.

“The AI Optimal Policy Algorithm functions like a seasoned network architect, analysing traffic patterns and application requirements to automatically generate optimised network configurations and security policies,” he adds. “For heterogeneous networks, it intelligently adapts to the unique characteristics of multi-vendor devices, harmonizing their parameter settings to ensure consistency.”

Huawei has designed the solution to meet the reliability requirements of financial services institutions. The company says that if a link fails, its M-LAG technology enables link failover in less than 100 milliseconds (ms). If a device fails, device-level failover is achieved in less than a millisecond. “These transitions occur so rapidly that applications and users experience virtually no perceptible disruption,” says Mr Wang.

Enhancing network security

Financial services institutions need highly secure networks to protect assets and customers. In the event of an attack, the potential loss is huge.

The Xinghe solution uses automation and AI to improve security. “Traditionally, network and security teams had to collaborate manually to configure security policies on each device – a process that was not only time-consuming but also prone to errors,” says Mr Wang. “Now, the NCE platform autonomously analyses application access requirements and generates optimised security configurations, significantly reducing the operational burden on administrators.

“The common encryption algorithm on the network (MACSec) can encrypt only 78% of core data,” adds Mr Wang. “The data address and sending frequency are still leaked and could be used by criminals.”

He explains: “Our physical layer encryption encrypts the OSI physical layer (the raw bit stream). This is faster and more comprehensive than traditional high-level encryption because it protects all data flowing through the network, including protocol header information that is not normally encrypted. This provides comprehensive data protection, while the encryption process has almost zero impact on performance due to its hardware-level implementation.”

Eliminating the trade-off

“In the past, companies had to make a trade-off between unity and diversity, either accepting the limitations of a single vendor or taking on the complex management of a multivendor environment,” he says. “Our solution breaks this dilemma by making multi-vendor networks as easy to integrate as a single system while retaining the flexibility to select the best technology from different vendors.

“We’ve targeted financial datacentres with our solution because financial services institutions require both exceptionally high reliability (to meet regulatory requirements) and extremely low latency (to fulfil business demands), while often operating within complex multi-vendor IT environments,” says Mr Wang. “If our solution can meet the stringent standards of the financial services industry, it will naturally be adaptable to the needs of other sectors.”