Yankuang Group

Business Background

The Yankuang Big Data Engineering Project primarily focuses on the implementation of an ERP system, aiming to leverage the horizontal integration and vertical penetration advantages of the ERP system. The goal is to achieve centralized and standardized management of core business operations such as personnel, finance, and materials, supporting the realization of Yankuang’s strategic objectives, implementing the group’s integrated management model, and enhancing the company’s operational management level.

As the foundational IaaS cloud platform faces the following challenges:

1. The group’s big data engineering business systems are numerous, leading to an increased demand for IaaS resources, and the current infrastructure is insufficient to meet the growing business needs.
2. The big data engineering project uses the dice one-stop PaaS platform for system construction, which requires higher flexibility, agility, and openness from the IaaS platform.
3. The PaaS platform demands higher standards for resource orchestration, rapid delivery, and automated operation and maintenance management of the underlying infrastructure.

Based on these requirements, the client needs to build a private cloud platform, pooling hardware resources, and improving equipment resource utilization through on-demand allocation, dynamic scheduling, and other methods to respond to business needs.

Solution Deployment

Through the construction of two phases of the project, an integrated cloud platform from IaaS to PaaS and SaaS has been built in Yankuang Group’s data center, supporting the Yankuang Big Data Project.

• Phase 1: The project utilized local storage, adopting ZStack Cloud virtualization technology to pool service resources, allocate on-demand, and enable elastic computing. This approach fully utilizes hardware resources and establishes a cloud host environment for the dice basic R&D and testing platform. The one-stop PaaS platform was also built to support the DevOps full-link technology, creating a closed-loop management system from code construction, continuous integration, continuous deployment and delivery, to operation and maintenance monitoring. This laid a solid foundation for the Yankuang Big Data Project.
• Phase 2: As the Yankuang Big Data Platform continued to develop, business applications based on the dice platform were constantly being launched, requiring the expansion of the underlying IaaS to support the growing resource demands. The Phase 2 project expanded the chassis, and the main storage adopted FC-SAN centralized storage. The cloud platform was divided into two major clusters. The business systems in the original local storage cluster were migrated to the newly built high-performance centralized storage cluster in Phase 2. The original local storage cluster was used as the Yankuang Big Data Engineering development and testing platform, supporting the development and testing of the group’s business systems. The new high-performance cluster in Phase 2, through the cloud platform and centralized storage, provided high availability (HA) and high IO capabilities for business systems, meeting the growing production business needs of the group.

Project Benefits

1. Improved Server Resource Utilization through Resource Pooling ZStack Cloud’s virtualization technology allowed the project’s servers to be virtualized and doubled, meeting the early-stage development and testing needs of the big data project, enabling rapid startup and deployment.
2. Hierarchical Resource Environment Allocation to Improve Service Quality The cluster functionality of ZStack Cloud allows logical division of resources into different performance levels. In this project, the local cluster and centralized cluster supported different business applications. High-performance resources and business high-availability services were provided for critical business applications through centralized storage clusters, while more economical resource environments were offered for less critical applications using local storage, improving service quality and resource allocation.
3. Enhanced Unified Management through Multi-Tenant and Hybrid Cloud ZStack Cloud’s multi-tenant feature allows the configuration of different accounts and quotas to meet the resource demands of different software vendors. The hybrid cloud functionality seamlessly integrates with Alibaba Cloud’s public cloud services, enabling unified management of both online and offline resources.
4. Faster Business Deployment through Flexible Cloud Platform Management ZStack Cloud’s open API interfaces were integrated with the dice platform, and functionalities such as resource orchestration supported the rapid release and deployment of dice platform applications, improving business delivery efficiency.