This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Search

  • banner pc

    Accelerating the Transition to Financial Clouds

To retain customers, financial institutions must meet increasingly higher requirements for services. Transitioning to cloud computing offers bankers a way to become more competitive while improving their efficiency.

Competition in the banking business is fierce, and nowhere more so than in China. A few months ago, for example, China’s Ant Financial Services Group, an affiliate company of eCommerce giant Alibaba Group Holding Ltd., closed the world’s largest round of private equity financing for an Internet company at USD 4.5 billion (CNY 30.9 billion) that gave the company a post-money valuation of roughly USD 60 billion (CNY 412 billion).

The China Banking Regulatory Commission recently began to explore the construction of public financial clouds for the banking industry as well as hybrid clouds that would combine private and sector-specific applications. The commission suggests that banks implement cloud computing architectures that decouple applications and infrastructure, and is encouraging banks to migrate their Internet-oriented information systems to cloud computing platforms by the end of 2020.

Huawei views this migration as three stages that will gradually increase the use of cloud resources.

New Clouds on the Rise

Financial clouds represent a business approach in which specific data, processes, and services are moved to cloud computing platforms. The major objectives of financial clouds include:

  • Improved user experience
  • Enhanced system agility
  • Greater computational capacity
  • Restructuring of data value

The goal is to enable financial institutions to provide improved services to their customers while reducing Operating Expenditures (OPEX).

Private clouds are built for internal purposes, self-owned branches, and key services. Early adopters have begun to lease surplus resources and services to other financial institutions, branches, and upstream and downstream partners. Internet-based services that have bursty loads or less stringent security requirements are candidates for being reassigned to a public cloud.

For cloud computing or any other Information Technology (IT) architecture, managers of mission-critical systems must choose servers, software, and storage that suit reliability and performance goals. Around the world, managers are rethinking traditional choices. The trend in China is that businesses are actively seeking to reduce their reliance on the so-called ‘IOE’ system. IOE stands for IT network components based on the names of three suppliers: IBM, whose servers are packaged with the UNIX operating system; Oracle, the database management system vendor; and EMC, a supplier of data storage hardware. Dubbed the ‘De-IOE Movement,’ this effort has been ongoing for many years and is a key motivator in the drive toward the wider adoption of cloud infrastructures.

A qualifying factor for the movement of banks to replace UNIX-based servers is the widespread availability of open platforms that are based on x86 servers and Virtual Machines (VMs). The reality is that many x86 servers do not provide the performance, reliability, and availability necessary to match the long-proven, mission-critical reliability of the UNIX platform. Just one example is the inability of x86 servers to support component replacements without service interruption.

Huawei’s KunLun 9032 x86 Mission-Critical Server is designed to accommodate the replacement of CPU and memory modules as well as other core components without interrupting ongoing services. With up to 32 CPUs and 576 cores, a single KunLun 9032 delivers levels of performance that rival mainstream UNIX servers. When deployed in conjunction with Huawei’s OceanStor 18000 V3 high-end storage, the KunLun 9032 meets the high-performance and low-latency requirements for efficient and trustworthy banking services.

Financial clouds for specific data, processes, and services

From Virtualization to Cloudification

Many banks around the world necessitate that all new applications be deployed on VMs. Huawei refers to this stage as Financial Cloud 1.0. While VMs are certainly useful, they apply only to the layer where computing nodes reside. This limitation raises questions concerning network and storage devices, and how managers will efficiently configure thousands of service policies and deal with the databases on physical nodes.

To help address these concerns, Huawei’s Financial Cloud solution employs enhanced OpenStack-enabled Ironic components while leveraging Huawei’s extensive experience in Software-Defined Networking (SDN) and distributed storage. The Ironic-based solution installs management software for network and storage automation onto a ‘bare-metal’ architecture that enables banks to roll out new services quickly without having to modify databases manually or write scripts.

Stepping up to Financial Cloud 2.0

Cloud management platforms are the most important core element of the stage that Huawei calls Financial Cloud 2.0 — and basic virtualization and resource provisioning are not enough. Modern financial institutions need unified pools of computing, network, and storage resources that enable collaboration across business applications. Unified resource pools are coordinated among multiple data centers to instantiate business-driven automated management and scheduling.

To implement these platforms, financial institutions choose from three main approaches:

  • Undertaking in-house, open-source software development projects. This is a short-term approach that features low costs on a small scale.
  • Purchasing proprietary, commercial off-the-shelf software. Although this path features mature applications, the trade-offs are high upfront costs and poor compatibility with other software.
  • Purchasing commercial open-source software, which also features mature applications and is standards compliant (highly compatible) but requires long-term planning.

Traditional financial institutions prefer the third approach, particularly when operating within the OpenStack architecture. As a de facto standard for private clouds, OpenStack has matured rapidly. Fortune 500 companies, such as BMW, Disney, and Walmart, are all using OpenStack in their production environments. The number of financial institutions deploying OpenStack is increasing year by year.

Huawei provides solutions that both enhance OpenStack and decouple OpenStack modules. The ability to decouple modules from the software stack allows components to be upgraded individually, without impacting services. On the order of 200 Linux security patches are released each year, and, generally, it is expected that entire VM arrays will need to be restarted for each patch to take effect. Cloud industry statistics reveal that, on average, more than 70 percent of hot patches require VM restart. Amazon Web Services (AWS) registers at about 50 percent. By comparison, Huawei supports 75 percent of the patches without VM restart, which significantly helps banks that employ 2,000 to 10,000 VMs.

Disaster recovery and security compliance are two other critical aspects of financial IT. For mission-critical services such as online banking, the Recovery Time Objective (RTO) must be less than six hours, and the Recovery Point Objective (RPO) must be less than 15 minutes. To confidently migrate their operations to the cloud, banks need automatic emergency handover, Backup-as-a-Service (BaaS), and other powerful functions.

In 2015, Huawei initiated an OpenStack project to provide users with data protection orchestration — or, Data Protection-as-a-Service (DPaaS) — for applications, VMs, and network resources in the cloud environment. Now called Karbor, this project’s software integrates various backup software packages to provide back-end, enterprise-grade backup and recovery capabilities.

The Next Stage

Financial institutions need to provide users with mobile Apps and services designed to support distributed deployment and are extremely fast and offer flexible scalability. Furthermore, these institutions must consider the construction of platform capabilities based on Infrastructure-as-a-Service (IaaS) so they can introduce quicker iteration capabilities to the full App lifecycle (development, testing, releases, and upgrades).

Huawei calls this stage Financial Cloud 3.0, where the container-driven, lightweight Platform-as-a-Service (PaaS) model is a key technology. PaaS is especially applicable to newly constructed cloud native application systems. In addition to supporting physical and virtual resources, IT infrastructure resource pools must support unified deployment and management of container resources.

Based on the industry’s mainstream open-source architectures, such as Docker and K8S, Huawei’s Financial Cloud solution provides a number of enhancements and advanced functions that fully leverage the advantages of lightweight containers. In the development and testing environments of banks, the solution has improved the application deployment density by four times and reduces the amount of time taken to deploy individual applications from minutes to seconds.

For financial institutions that must release version updates periodically, this functionality is especially useful in the development and testing period. Additionally, Huawei offers the FusionStage PaaS platform, which gives enterprises a complete environment for App development, deployment, operation, and management. This open architecture provides a wide range of middleware services, including those for Big Data and the IoT as well as application components specific to different types of enterprises. FusionStage supports mainstream development languages and protocols, and allows customers to deploy millions of containers as well as new microservices in less than a minute.

Huawei has implemented container-based I and P layers to implement full-process automation of IBM WebSphere® Application Server (WAS) Apps across regions and data centers, improving IT resource utilization and service agility. By introducing distributed Server Load Balancers (SLBs) that support the gray release function, Huawei has achieved access for global routes.

Unified Asset and Resource Management

The Huawei Financial Cloud solution uses a cloud platform architecture to manage self-help service directories and resource pools for web and App resources, transaction data, Big Data analytics, and virtual desktop infrastructure. The solution allows users to combine their own private clouds with managed cloud services to implement unified management and service-oriented utilization of their own and third-party data assets and resources.

The solution also enables the implementation of financial IT infrastructure platforms that support smooth evolution based on IaaS and PaaS resources provided by private clouds. Huawei’s solution includes reliable hardware, a cloud operating system, cloud management platform, and PaaS platform. At the same time, the solution’s open-source architectures and standard Application Program Interfaces (APIs) support third-party PaaS and are compatible with native standard northbound and southbound interfaces.

Thanks to ecosystem-friendly features, Huawei’s solution has attracted contributions and support from more than 150 financial software vendors globally, including India’s Infosys, France’s Sopra Banking Software, and China’s Yucheng Technologies and ChinaSoft International. These firms collaborate with Huawei to develop innovative applications and provide leading solutions to over 300 financial customers worldwide.