Lian Life IFRS17 Project

1. Background

The new insurance contract standards (i.e., CAS25 or IFRS17) involve massive data processing, entirely new actuarial and accounting treatment, and brand-new systems and process support. Data is fed from the insurer's legacy core business systems into Jinke's integrated IFRS17 product for processing and refinement, and then pushed from the subledger back to the general ledger system to complete financial statement disclosures, aligning with the new standard requirements. Lian Life Insurance Co., Ltd. is a nationwide life insurer that previously lacked a distributed database capable of supporting IFRS17 data integration and analysis. By upgrading its systems with GBase 8a MPP Database running on an open x86 architecture, Lian Life significantly enhanced its big data computing power while keeping costs under control, meeting period‑end closing timeliness requirements.

2. Overall Objective

Select GBase 8a MPP Database as the distributed database platform for the IFRS17 implementation project. It will provide the essential data integration and analysis capabilities for IFRS17, serve as the data warehouse foundation for the new insurance contract standards system, and support the generation of accounting reports compliant with the new standards.

3. Requirements

1) The system must employ a fully parallel MPP distributed architecture, with no single point of contention; avoid single-point performance bottlenecks and single points of failure (SPOF); all nodes are shared-nothing with symmetric computing capability.

2) Support 64‑bit Linux operating systems such as CentOS, Red Hat, SUSE, and Kylin.

3) Support SQL 2003 OLAP functions and programming interfaces including C API, Python API, and TCL API.

4) Support tiered data storage across traditional databases, MPP databases, and Hadoop platforms based on data lifecycle theory. Users can read and write transparently through SQL, achieving unified access to data at different lifecycle stages.

5) Provide enterprise management tools and cluster monitoring tools that enable centralized management of all database functions, as well as monitoring and automated tuning of system running status, resource usage, and task execution.

6) Support parallel data loading and exporting from multiple sources, including various network protocols such as ftp, sftp, hdfs, and http; support real‑time data loading from Kafka; support parallel loading of a single table from multiple data sources.

7) Non‑exclusive reads and writes: data can be loaded and queried concurrently, with concurrency capacity exceeding 1000.

8) Support DBLink functionality between homogeneous and heterogeneous clusters, including write operations on DBlink tables. For homogeneous clusters, single‑node data transfer performance must be at least 100 MB/s.

9) Capable of importing data from and exporting data to Hadoop platforms, enabling backup and recovery of massive data volumes based on Hadoop.

10) Support distributed shard storage by node, hash‑based shard storage, and multi‑column hash distribution strategies.

The IFRS 17 product is a comprehensive system, distinguished by its powerful calculation engine. In line with IFRS 17 accounting standards, it processes large volumes of raw insurance business data from upstream systems, applies complex, rigorous, and interdependent formulas along with multi-dimensional actuarial assumptions, and performs measurement and estimation across various levels. This generates vast amounts of intermediate data, which is ultimately aggregated into outputs that comply with IFRS 17 specifications. The IFRS 17 platform supports multi-source, real-time, and efficient data integration, and delivers vectorized computation, hardware acceleration, and stream-batch unification to meet real-time data processing demands.

This project uses an 8-node GBase 8a cluster as the data foundation for the IFRS 17 solution, enabling the system to efficiently handle IFRS 17 data and business analytics. GBase 8a, a shared-nothing MPP database, leverages its distributed architecture to provide robust support for IFRS 17 under high concurrency and massive data volumes, ensuring excellent scalability, stability, and high performance.

IFRS17 Business Support: GBase 8a enables batch job processing, model computation, real-time computing, stream computing, and interactive queries.

High Performance: GBase 8a boosts data storage and computing capabilities for the IFRS17 platform, shortens batch processing times, supports real-time computing, and unifies data storage and sharing to further reduce storage and ETL costs.

Big Data Platform Integration: It establishes a one-stop closed-loop process covering data integration, development, governance, services, and asset management, standardizing the entire data development workflow and improving the efficiency of data flow across the IFRS17 platform.

Independent and Controllable: GBase 8a meets the requirements of China’s IT Application Innovation (Xinchuang) ecosystem, providing online scalability and high availability to safeguard business growth and continuity.