Big Data Solutions

QingCloud provides mainstream big data components that cover data collection, storage, processing, analysis, and display. In this connection, we provide enterprises with basic big data service resources with high performance and reliability, assisting them in driving service decision and product intelligentization.

Business Challenges


Support for Ultra-High Concurrency
Support flexible online scaling in real time without the need for data migration; help enterprises to easily cope with performance requirements for ultra-large capacity and ultra-high concurrency.
Data Security and Reliability
Provide secure, reliable, and low-cost storage services on the cloud, with the data persistence reaching up to 99.999999999% and service availability reaching 99.99%.
High Resource Utilization
Use an architecture with separated compute and storage, improving the resource usage by 50% and greatly reducing service resource costs compared with the coupled architecture.
Open Cloud Ecosystem
Provide customers with third-party tools and components for machine learning and algorithm modeling, so as to deliver comprehensive big data solutions.

Solution Architectures

Architecture characteristics
  • Provide big data services on the cloud based on the Linux container (LXC), avoiding the IO loss of KVM, and achieving almost the same performance as that of bare metal servers.
  • Support receiving, storage, and conversion of data with multiple types, formats, and modes on the same platform, solving the data complexity issue.
  • Support flexible online scaling in real time based on service needs, improving service processing capabilities.
Scheme effect
  • While meeting the high-performance requirements of customers for big data business, we can reduce the cost of server resource input and effectively improve resource utilization.
  • Provides intuitive and easy-to-use one-click operation, and the deployment of a big data cluster can be completed in minutes, and the operation is easy to get to work.
  • Maximize resource utilization, use resources on demand, and meet the high-growth demand of business volume.
Architecture characteristics
  • HashData uses an architecture with separated compute and storage. Data is stored on the QingCloud Object Storage, while the compute layer uses the MPP compute engine based on the Greenplum kernel.
  • The system supports unlimited horizontal scaling, with linear improvement of performance in data access.
  • Hot and cold data management functions provided by HashData ensure data query performance.
Scheme effect

1. Compared with the architecture with coupled compute and storage, separation between them ensures query performance, greatly reducing server resource costs.
2. Secure and reliable storage services on the cloud enable data persistence to reach up to 99.999999999% and service availability to reach 99.99%.
3. Only 1/10 of storage costs for the traditional solution are required while the data query performance is ensured, greatly reducing IT costs.