Currently, enterprise applications obviously tend to be mobile and intelligent and the types of the data generated by applications are changing, with pictures, videos, voice, documents and other unstructured data absolutely dominant in the total enterprise data. Moreover, traditional storage has become a bottleneck in the digital transformation of enterprises and it is a must to choose more appropriate storage technology for IT infrastructure construction.
At present, more than 80%of new data are unstructured data;
More than 95%data is unstructured data in the future.
Source: IDC & Inspur - Report 2019 on data and storage development research; Huawei GIV @ 2025
Why object storage is needed?
The flexibility and flatness of object storage make it possible to store a large amount of data by means of expansion, save enough information via metadata to help applications find data as quickly as possible and be excellent in unstructured data storage and utilization. Furthermore, its large-scale expansion and storage of any type of data in a native format economically and efficiently has made the object storage technology quickly become an efficient information processing platform for data owners.
SandStone MOS is an object storage product both competent in enterprise-level storage and intelligent retrieval capability. The new generation of storage engines has built a cross-regional storage architecture based on standard commercial server, which can be linearly expanded and has almost no limit to the maximum capacity. Besides, while providing high reliability and high availability, it has integrated the capabilities for data intelligent processing and analysis, simplified the infrastructure needed for massive data processing, and thus greatly improved the efficiency of data processing.
Linear scale-out based of general hardware, support storage of 10 billion files
Flexible and efficient multi-center disaster recovery scheme
Data lifetime automation management, seamless integration with public cloud
Intelligent analysis/data value mining
Enterprise content management
- Support Hadoop, Spark and other big data analysis platforms and provide
- complete solutions for collection, storage and analysis of different business data, such as Cloudera big data platform, telecom operator log analysis, financial fraud detection and gene data sequencing.
- Provide a highly extensible data carrier platform for the massive data from applications of face recognition, speech recognition, natural language processing, machine learning and other AI technologies in mobile terminals, UAV equipment, vehicle intelligent equipment, remote business terminals and other scenarios; moreover, create a thorough solution for data analysis, value feature extraction and information retrieval based on AI technologies.
- Some documents are huge in number and need long-term storage, such as bank bill images, audio/video records, face images, paper document scanning copies, insurance company's customer service voice data, telemarketing voice data and hospital medical images. Almost infinite scalability of object storage can satisfy the demands of increasingly growing data and the access between storage devices and services is simple and easy to maintain, so object storage can be used as a data source to directly analyze big data and mine the valuable information in image data.
- According to the access frequency, object storage system can migrate data automatically in multiple storage resource pools (hot, cold and cloud), combine the accessibility of near-line storage with the economy of cold archiving, and provide long-term data protection, nearly unlimited scalability and very low overall cost, which can meet the needs of massive data archiving, production team collaborative access and disaster recovery.
Big data storage and analysis
- Support Hadoop, Spark and other big data analysis platforms and provide complete solutions for collection, storage and analysis of different business data, such as Cloudera big data platform, telecom operator log analysis, financial fraud detection and gene data sequencing.
- Object storage system can provide persistent, safe and cost-effective cloud storage capacities, thus to meet various needs for backup. Back up data from data center to object storage system, which can replace expensive and inefficient tapes, and simplify backup operation and archiving. For the data in the cloud, the object storage system can be used to realize consistent automatic backup and recovery solutions by virtue of the cloud native interface and low cost.