Sichuan Rural Commercial Bank Powers Down Mainframe, Rebuilds Core on Alibaba Cloud

1. Decision: From Closed Mainframe to Open Cloud Architecture

In 2017, senior staff at Sichuan Rural Commercial Bank raised the provocative question of whether a mainframe could support the bank’s next twenty years, confronting the high cost, rigidity, and scaling limits of traditional mainframe systems. With only eleven Chinese banks still operating large mainframes, the bank faced escalating maintenance costs, limited extensibility, and regulatory mandates for domestic‑controlled technology. Recognizing the need for a sustainable, scalable foundation, the bank decided to pursue a distributed cloud‑native architecture despite the lack of production‑grade examples at the time.

2. Reconstruction: Rewriting Code and Architecture

In December 2025, project leaders from the bank, Alibaba Cloud and partner Zhongdian Jinxin signed the mainframe shutdown ceremony, symbolizing a commitment to a multi‑year rewrite rather than a simple lift‑and‑shift. The initiative was organized as a three‑stage, two‑step program: planning and platform construction, middle‑platform development, and full‑scale application migration. Alibaba Cloud supplied the cloud‑native foundation and engineering tools, while Zhongdian Jinxin executed the core‑system code refactoring. Every business module—credit, payments, accounts, and transaction logic—was decomposed into independent services, re‑engineered for distributed deployment, and repeatedly validated through performance testing, disaster‑recovery drills, and production‑level roll‑backs.

3. Leap: System Capability After Migration

Post‑migration, the ShuXin Cloud platform now serves nearly 70 million customers, over a hundred million accounts and processes tens of millions of transactions daily. The multi‑active, cross‑data‑center design achieves seamless failover within five minutes with zero data loss and zero accounting errors. Core online processing capacity has increased by more than three times, concurrent transaction throughput has risen from the low‑thousands to multi‑thousands per second, and end‑of‑day batch processing is about 38 % faster. These technical gains translate into faster account opening, reduced loan approval cycles from days to minutes, and robust 24/7 intelligent chatbot services.

4. Conclusion: A Proven Path for Future Banking

The eight‑year transformation demonstrates that a full rewrite to cloud‑native architecture can deliver orders‑of‑magnitude improvements in resilience, scalability and operational efficiency, while also laying a “growth‑ready” foundation for AI‑native financial services. The project’s documented metrics and engineering practices provide a replicable blueprint for other regional and cooperative banks seeking to replace legacy mainframes with modern, self‑evolving cloud platforms.