Project background

Through a long-term planning, Protein Scientific Research Center (Shanghai), Shanghai Institute for Biological Sciences of Chinese Academy of Sciences proposed to built a dedicated database and computing analysis comprehensive service system for the protein facilities. Through developing and determining the data software system for laboratory data acquisition, analysis and integration, strive to form an international advanced national protein scientific research comprehensive database, establish and development high-performance computing and system biological technical conditions platform, constitute a complete information processing and computing center through the two parts above, thus to build a mathematical model for the complex biological system, implements studies on high-efficient protein dynamics, interaction prediction, molecular design, complex biological system simulation, etc. So that to provide powerful database and computing analysis abilities guarantee for relevant facilities in Shanghai.

Solution introduction

To build a better and effective information infrastructure for the data center, improve the utilization of the application systems in the data center, achieve system data resources sharing to the largest extent, Sugon designed the highly-reliably data storage and backup system for the entire protein scientific research platform, as shown in the figure below:

The storage resources respectively adopt primary storage (online storage) and secondary storage. The primary storage applies fiber-optic high-speed disk and is equipped with high-cache controller, it provides databases, cloud platforms and other applications with higher I/O reading and writing requirements. In addition to the fiber high-speed storage system, in view of the high-performance computing analysis platform, a set of high speed Parastor parallel storage system was arranged for storage of unstructured data. The secondary storage applies the high cost-efficient SATA disk array and is equipped with Infiniband interfaces, thus the fundamental data sharing and offline backup demands may be satisfied.

The backup system adopts Sugon DBSTOR backup storage system and LTO5 high-speed taps library package. Sugon DBSTOR is an enterprise backup storage system, it supports data off-site disaster recovery and backup system auditing. The taps library is equipped with 250 LTO5 high-speed taps, the number may be increased according to business developments. Data backup of critical core data may be completed within 20 hours, for non-critical core data, backup may be completely finished within 72 hours. The system remote disaster recovery function may be achieved through the taps off-site storage.

Project summary

This solutions designed a multiple-level storage backup system according to the database and computing analysis platform. When a SAN+NAS network storage system is built, the initial storage bare capacity is close to a PB, the effective capacity may be expanded to PB-level data disk array storage without any influence on its performance. Various platforms may store computing data fastly and reliably, so that to achieve high I/O and large throughput data access performance, in this solution, remote disaster recovery module is reserved. Therefore, this solution is able to meet the system data storage demands, achieve data centralized storage, information sharing, centralized backup and automatic backup.