Full coverage for finance scenarios
Payments and transfers
Finance scenarios such as payments, transfers, and accounting demand high throughput. The SOFAStack distributed transaction product has shown excellent performance during major sales promotions, demonstrating that performance is not a bottleneck.
Wealth management
These scenarios often involve large sums of money and require high product stability. The SOFAStack distributed transaction product offers financial-grade quality to ensure business continuity and stability.
Insurance and regulatory reporting
These business types are typically complex and involve multiple parties. The SOFAStack distributed transaction product has evolved for over a decade. It is flexible enough to handle various scenarios, meet transactional consistency requirements, and integrate seamlessly with your business.
Simplified payments for government services
Bill payments
Bill payments are an essential part of government services and an extension of payment and transfer scenarios. People pay for utilities, phone services, and internet access using mobile apps or computers. Government systems must process this payment information consistently. The SOFAStack distributed transaction product ensures that related information is updated in sync and that data is promptly synchronized across systems.
Instant cross-region information sync
Government agencies in different regions often maintain their own local databases for initial registrations, such as personnel changes and business ICP filings. In an age of frequent information updates, manual synchronization can cause dirty reads and writes. The SOFAStack distributed transaction product ensures that government data is synchronized efficiently and accurately.
Versatile scenarios across the Internet
Orders, membership cards, growth points, and loyalty points
In a points mall, for example, purchasing a product with a membership card balance involves multiple services. These services include deducting the account balance, adding loyalty points, increasing growth points, and updating the order history. The traditional reconciliation method for these scenarios is slow and costly to adapt to business changes. You can simply integrate the SOFAStack distributed transaction product to synchronize this data.
Escrow transactions
Consider a payment scenario for an E-commerce flash sale, which requires high performance to handle high concurrency. The typical flow is to deduct the user's funds, freeze them, and place them in an intermediary staging account. In an escrow transaction, the money is not sent to the merchant immediately. After a set period, such as seven days, the funds are transferred from the intermediary account to the merchant. The SOFAStack distributed transaction product supports large-scale flash sales and ensures successful payments. During off-peak hours, it can process the payment data and asynchronously execute the fund transfer. This ensures the funds are successfully transferred to the merchant's account.
Financial-grade quality assurance
Financial-grade disaster recovery
The product provides multiple modes and levels of disaster recovery, including intra-city and cross-region options. It adheres to the highest industry standards to ensure your business continuity.
Unparalleled performance
Compared to the traditional two-phase commit pattern, the product reduces lock holding time to significantly improve performance. Its unique Performance Bursting Mode greatly increases throughput. It supported 256,000 payment operations per second during the 2021 Double 11 shopping festival.
Simple and easy to integrate
Based on years of practical experience at Ant Group, the product is fast and flexible to integrate. It is also easy to use and maintain.
Guaranteed compatibility
A distributed transaction is an abstract concept at the Service layer, independent of the underlying transaction implementation. This means you can integrate various data stores into a distributed transaction. These include relational databases such as MySQL and Oracle, key-value (KV) stores such as MemCache, and column store databases such as HBase. You can simply wrap their operations as participants in the distributed transaction.