Core Data Warehouse and Data Mart Construction Project

The rapid evolution of mobile internet and smart devices poses significant challenges for the telecom industry:

First, shifting market demands. Customer needs are increasingly diverse and niche, with an explosion in product variety and expanding touchpoints. Traditional mass-marketing models can no longer keep pace with this new reality.

Second, intensified competition. The mobile internet era has spurred rivalry not only from within the telecom sector but also from adjacent industries. As internet companies surge, telecom operators are losing their once-dominant position in the value chain, and the evolving ecosystem fuels fiercer competition.

Finally, technical challenges. The rapid growth of mobile internet generates massive data volumes for business support systems, with expanding data types that now include network data, transaction data, user profiles, and location signals. Traditional data processing technologies have become the bottleneck for system scalability.

To tackle these challenges, Zhejiang Mobile aims to seize big data opportunities, adapting to demand diversification, fragmented behaviors, and massive application loads. It is transforming its marketing mindset with a 'big data, hyper-segmentation, and micro-marketing' approach to revamp pricing, channels, marketing, and communications—building a service system purpose-built for the mobile internet era.

Zhejiang Mobile's enterprise digitalization initiative needs to support the go-live of 5–10 internal applications and 5–10 external applications, enabling data monetization and revenue generation;

Build an MPP resource pool cluster, consisting primarily of a core data warehouse and data marts;

The primary data warehouse MPP cluster mainly handles data modeling and foundational data storage and computing for the BSS (Business Support Systems) and OSS (Operations Support Systems) domains;

The data mart MPP cluster primarily supports internal applications and external monetization services. Traditional data mart applications such as VGOP, ESOP, and the Innovation App Incubation Platform are also being gradually migrated to this data mart MPP cluster.

Overall Architecture of the Big Data Foundation Platform (Current Phase):

All data exchange between the MPP cluster and external systems is handled through the cloud-based ETL platform. Application development for the MPP cluster is based on DACP (Data Management Subsystem). The cloud management execution center collects MPP cluster metrics and provides them to the cloud resource management platform;

The data warehouse MPP cluster is primarily responsible for data modeling and foundational data storage. After the new business goes live, the legacy data warehouse services will be decommissioned. The GBase 8a MPP Cluster database platform uses a 3+61 architecture, with 3 management nodes and 61 data nodes, and the current data volume is 732 TB. Within the cluster, a 10 Gigabit Ethernet network provides high-speed connectivity;

The data mart MPP cluster mainly hosts internal applications and external monetization applications. The GBase 8a MPP Cluster database platform uses a 5+62 architecture, with 5 management nodes and 62 data nodes, and the current data volume is 620 TB.

Low Cost: Reduces overall system cost to approximately 1/10 of the original;

Elastic Scalability: The GBase 8a MPP Cluster database efficiently handles petabyte-scale data, meeting storage needs for foundational and application data. It shifts from traditional vertical scaling to horizontal scaling based on data volume, enabling dynamic expansion without service interruption and ensuring business continuity.

High Availability: Achieves effective load balancing through proper configuration, fully leveraging the computing power of each node to enhance overall system collaboration efficiency. A redundant backup strategy ensures service continuity even if a node fails.

High Performance: Boosts data service capabilities to enable data monetization and generate revenue. Delivers efficient data processing and query performance for various specialized analyses and innovative application requirements.