Catastrophe Risk Management Platform (Phase II) Construction Project

China RE Catastrophe Risk Management Company LTD. (“China RE Catastrophe”) is China's first fintech company focused exclusively on catastrophe risk management, jointly established by China Reinsurance (Group) Corporation and multiple institutions . As digital transformation deepens across industries, enterprise business scenarios are becoming increasingly diverse, high-concurrency, and high-velocity. Business data volumes continue to grow explosively, imposing extremely high demands on data storage clusters in terms of flexible adaptation, elastic scaling, concurrent processing, and disaster recovery availability.

Focusing on core business pain points, the company optimized four core capabilities — data distribution, cluster scaling, concurrent processing, and data disaster recovery — to advance the China Catastrophe Risk Map Platform (Phase II) construction. The system abandons traditional rigid cluster operating models in favor of customizable data distribution strategies, online transparent elastic scaling, and high-concurrency parallel read/write processing. It also establishes a multi-tier data redundancy and primary-standby cluster disaster recovery framework, comprehensively resolving the performance bottlenecks, insufficient flexibility, and weak availability of traditional data clusters. The platform perfectly adapts to enterprise business scenarios of varying scale and type, significantly reducing operational costs and business interruption risks while ensuring efficient business operations — providing solid underlying data support for stable deployment and efficient iteration of the enterprise's digital business.

The system is built on GBASE's GBase 8s database. A 3-node cluster supports thousands-level concurrency, ensuring stable operation during business peaks. It supports online cluster scaling with zero business impact and scaling speeds of up to 20 TB/hour, dynamically adapting to data and traffic fluctuations and allocating resources efficiently. It supports two data distribution modes — HASH and RANDOM — configurable on demand to avoid data skew, adapt to various business scenarios, and optimize overall cluster performance. It supports customizable 1–2 replica redundancy with automatic fault detection, data synchronization, and self-healing capabilities. Failures do not affect business operations, and no manual intervention is required. It supports primary-standby cluster deployment, full/incremental data synchronization, built-in fault rollback and recovery, and same-city disaster recovery — ensuring continuous business operation.

• Customizable data distribution: Avoids data skew, adapts to various business changes, and effectively improves resource utilization.

• Online transparent scaling: Rapidly responds to business fluctuations without downtime, reducing operational and business risks.

• Non-blocking reads and writes: Supports high-concurrency access, reduces access latency, and ensures stable, smooth operation during peak periods.

• Multi-replica: Multi-replica redundancy with automatic fault self-healing prevents data loss without manual intervention, significantly improving cluster stability.

• Disaster recovery: Primary-standby clusters with same-city disaster recovery enable rapid recovery from anomalies, effectively defending against extreme failures and ensuring continuous business operation.