OceanStor Pacific Scale-Out Storage

Knowledge mining from massive amounts of unstructured data can accelerate human genomics research, improve weather forecasting accuracy, assist diagnosis and treatment, help accurately locate oil and gas in energy exploration, and accelerate the R&D of autonomous vehicles.

With the advancement and widespread use of data analytics technologies, mass unstructured data is taking center stage in supporting enterprise production and decision-making. This in turn requires data storage to deliver more elastic scalability, higher efficiency, and better reliability.

On-premises data is distributed on different types of storage devices from different vendors, resulting in numerous data silos that hinder data utilization. Cross-region data is distributed in different data centers and edges, resulting in data isolation, which hinders instant access and unified protection. In addition, more and more data is distributed in both on-premises data centers and cloud infrastructure, making data mobility and collaborative production difficult.

Complicated management is the primary challenge caused by exponential data growth. Storage systems must be able to mask differences between heterogeneous devices, cross-region devices, and different IT infrastructures to simplify data management and improve data access efficiency.

Recent years have witnessed more frequent ransomware attacks on business systems. According to a survey, a ransomware attack occurred every 11 seconds in 2021, and the number of ransomware variants increased by 98% from 2021 to 2022.

In terms of ransomware protection, the network layer can only intercept known viruses in most cases. As the final carrier of data, storage systems are responsible for detecting attacks, retaining important data copies, and promptly restoring services to build the last line of defense for data security.

With the maturity of cloud native technologies and markets, more and more enterprises adopt multi-cloud IT construction to fully leverage the technical advantages of different cloud platforms. However, this practice has caused cloud data silos, preventing data from being shared between multi-cloud applications. Moreover, the services of different cloud platforms vary, hindering ecosystem integration.

Connecting one storage system to multiple clouds enables centralized data sharing, on-demand invoking, and multi-cloud application deployment, greatly reducing cross-cloud data migration and resource wastage while improving data movement efficiency.

As innovative services become production services and are continuously developed and segmented, there is an increasing gap between computing power and data lifecycles. Therefore, computing power and storage resources need to be flexibly and independently planned and maintained.

More and more enterprises use a decoupled compute-storage IT architecture and expect to build a highly reliable, cost-effective, and efficient solution for segmented scenarios through collaborative innovation of applications, computing, and storage.