Project background
The Big Data Center of Pudong New Area has the responsibilities of gathering data resources, providing a unified platform, and supporting rich applications.
Each bureau generates about 2PB of data every three years, and the amount of data stored online is increasing day by day.
The Data Storage Expansion Plan is started in the big data center to meet the data storage space of each bureau in the next five to 10 years.
Demand challenge
Traditional storage devices, such as tape libraries, CD-ROMs, and disk arrays, require manual data backup
It is easy to produce data silos, lack of unified management, and high cost of human maintenance and management
Facing the challenges of cloud storage separation, real-time call of data within three years, data availability from three to five years, and offline storage of data over five years.
Data layering controls costs and differentiates the frequency of data access
Scheme characteristics
UCloudStor unified distributed object storage was used to build 8PB data archive storage
Cold and hot data can be automatically transferred to the corresponding storage space. For example, cold data more than three years old can be automatically archived to the cold storage space
Fully distributed and redundant architecture, no single point of failure, multi-copy, EC, and fault self-healing data protection mechanisms
The three-node startup is lightweight and flexible, and provides log audit and refined access control policies, combined with WORM control and multi-form encryption, meeting data security and compliance requirements
Scheme value
A unified distributed storage system to solve the problem of long-term storage and management of massive data
Data is stored at different levels, enabling lifecycle management of production and archived data, simplifying development workload and human operation and maintenance costs
Data storage is highly reliable and available, maximizing data disaster recovery.
Convenient construction of smooth expansion, to meet the subsequent rapid landing, data bearing, on-demand expansion requirements
