GBASE Assists Shandong Mobile in Building City-level Data Marts

The city-level data mart clusters in Shandong had long relied on Oracle databases for support, resulting in persistent task delays and storage bottlenecks. Data flows between Oracle and AsiaInfo Hadoop also depended on application-side data processing and extraction, frequently causing data inconsistency issues that demanded an urgent resolution.

Using GBase UP to integrate Oracle, GBase 8a MPP, and AsiaInfo Hadoop in a converged architecture, efficient data exchange channels are built between engines through data virtualization and federation technologies. It combines computing models such as complex correlation analysis, stream processing, graph computing, and batch processing to deliver a big data platform that is unified externally and scalable internally. This allows massive data to be transferred from Oracle databases to MPP for processing, leverages MPP's high performance to meet the high-response demands of city-level data marts, and enables B and O domain data to flow through underlying database technology to Hadoop for in-database modeling and mining. A unified transparent interface is provided externally, so simple SQL can access and correlate data across engines. This streamlines data flows, greatly improving business responsiveness and supporting future growth.

1) GBase UP converges OLTP, OLAP, NoSQL, and other technologies to meet all data storage and computing needs, with built-in full data lifecycle management.

2) Achieve significant performance gains—data exchange between heterogeneous engines is automated at the underlying level, streamlining data flows.

3) Deliver external data services by leveraging the inherent advantages of local data marts, integrating unstructured data modeling to provide ready‑to‑use data models.

4) Enable deep, granular business analytics. High‑performance analytics empower customers to handle complex, real‑time, and performance‑intensive scenarios, efficiently manage massive data, and conduct multi‑dimensional, in‑depth analysis across all data types—accurately extracting valuable insights.