Unstructured data storage challenges
Unstructured data is growing at an annual rate of 30% to 60%. In response to the new challenges in enterprise it, software defined storage based on file system interface has become the first choice to solve the explosive growth of unstructured data in enterprise data centers. Nvmatrix provides hundreds of petabytes of storage space in a unified global namespace, supports tens of billions of files, and meets the needs of applications to access massive files.
Performance is critical
With the wide application of artificial intelligence, machine learning, gene analysis and high-performance computing, the requirements for the ability of high throughput and parallel access of distributed file system storage are becoming higher and higher. The performance of file system directly affects the efficiency of business operation. For applications such as artificial intelligence, it is even related to the core competitiveness of enterprises. Nvmatrix provides data support for core applications with high file system performance requirements.
Container persistent storage preferences
Kubernetes is the industry standard for container orchestration. All the most important business applications can be run on kubernetes. After business containerization, we are faced with the problem of how to access persistent data in the container environment. Traditional storage solutions face challenges such as interface adaptation and rapid data access after pod cross node reconstruction when serving container applications. Nvmatrix enters the container storage products of CNCF landscape to provide stable and reliable container storage services.
Product advantages
Global Namespace
- Single namespace: provides a unified, continuous, high-performance file global namespace, which is simple to use
- Rich protocol support: support NFS, smb/cifs, POSIX, FTP, CSI and other protocols
- Emerging business support: support emerging loads such as HPC, big data and containers
Flexible expansion
- Software definition, customizable node attributes, and support various brands of general x86/arm servers and domestic servers
- Flexible deployment, which can be extended to thousands of nodes to meet different business needs
- Expand on demand, and the performance and capacity will increase with the increase of node points to meet the performance and capacity requirements of growing businesses
- The whole pool is expanded, and the business is not perceived
Enterprise level functions
- Data redundancy: support multi replica and EC redundancy strategies, and provide multi fault domain management
- It supports file directory level snapshots, manual snapshots and scheduled snapshots
- Support DNS based load balancing and multiple load balancing strategies
- Support quota management, directory quota and user / user group quota. Quota includes capacity quota and file quota
Typical application scenarios
File sharing, enterprise office storage
Global namespace, easy to use. Support file sharing, network disk, FTP and other office scenarios.
Big data, HPC back-end storage
Compatible with HDFS and efficient file metadata processing mechanism, it can flexibly meet the requirements of ai/ml data analysis.
Video monitoring storage
Horizontal expansion, rolling upgrade, and permanent data storage.
Container shared storage
Support CSI interface, and support multiple pods to share data.