OceanStor 100D

Description

OceanStor 100D (formerly FusionStorage) is a fully self-developed intelligent distributed storage product that can be scaled out on a large scale. It can provide industry standard interfaces such as file storage, big data storage, and object storage for upper-level applications, eliminating the operational complexity caused by the construction of chimney-style storage systems, and helping enterprises to achieve more stable complex business loads, more efficient use of diverse data, and better storage costs for massive data.

• File storage: Compatible with native NFS and SMB protocols, compatible with POSIX and MPI, suitable for high-performance scenarios such as HPC (High-Performance Computing).

• Big Data Storage: Provides a big data storage and computing separation solution based on native HDFS, realizes on-demand configuration of storage and computing resources, provides a consistent user experience, and helps you reduce total cost of ownership; supports coexistence with the original computing and storage integrated architecture. Widely used in scenarios such as financial big data, Internet log retention big data, government big data, and smart city big data.

• Object storage: supports up to 100 billion objects in a single bucket without performance degradation, eliminating the trouble of bucketing and transformation of large applications; automatically grading to Blu-ray media, eliminating data migration and saving space. Widely used in production storage, backup or archiving of financial electronic bill images and dual recording (audio/video), medical imaging, government and enterprise electronic documents, and Internet of Vehicles scenarios.

• More stable carrying of complex services

End-to-end data integrity verification and self-healing; comprehensive sub-health detection and self-healing for hard disks, nodes, networks, etc.; support for simultaneous failure of 4 nodes or 4 cabinets without degradation of system reliability; end-to-end reliability assurance at the I/O level, device level, system level, and data center level, 24/7 service with zero interruption, providing high-availability storage services for complex enterprise applications

• Diversity data is used more efficiently

A series of technologies such as FlashLink® performance acceleration and AI Fabric lossless network ensure that the performance of 100 billion objects in a single bucket does not decrease, and massive small files can be easily accessed; multiple storage services such as files and objects are managed in an integrated manner, and data is used efficiently; new and old generations of hardware and multi-platform hardware coexist to achieve intelligent management of the system throughout its life cycle.

• Massive data storage costs are better

Massive on-demand expansion of 3-4096 nodes per cluster, linear growth of performance and capacity, and optimization of current investment; elastic Erasure Coding data redundancy protection, 91.6% hard disk space utilization rate*, 1.75 times higher than the traditional three-copy; rich main memory tiering strategies to achieve more reasonable storage and management of hot, warm, and cold data.