What are the security industry requirements for storage?

In recent years, with the popularity and gradual landing of the cloud concept of the security industry, cloud storage in video surveillance solutions at various project sites has become an indispensable part. It mainly includes the following aspects:

Large storage capacity and scalability

Monitoring points are built in stages and large-scale unified storage. The storage capacity is basically from 1P to 100PB+, and the general city-level storage scale can reach 100PB, which will reach a larger scale for first- and second-tier cities. Scalability is an essential part of the project, including storage capacity, storage performance, metadata management scale, and metadata access performance, all of which need to be scalable. Due to storage size and low video value, storage costs become sensitive in the industry, and the common three copies cannot be reached in the security industry. Therefore, high storage utilization is an essential part.

Hybrid storage, high performance

It satisfies the mixed storage of video and pictures, supports 7x24 hours of large file and small file writing, supports high-performance download of video, fast real-time on-demand and fast loading access of pictures.

High reliability, sustainable fault tolerance

When the number of nodes is satisfied, node-level fault tolerance can be achieved. When the number of nodes is not satisfied, disk-level fault tolerance can be achieved. It can dynamically sense the change of the number of nodes from one to many nodes and the number of nodes to one node, and dynamically realize adaptive fault tolerance. High-performance fault recovery quickly achieves data recovery without affecting business read and write, improving data reliability.

Easy to install, easy to monitor, maintenance free

Installation and deployment of minimal operations, expansion of one-click completion. Common disk failures, node failures, and single NIC failures can be completely free of real-time maintenance. Real-time monitoring of read/write traffic, read/write anomalies, faults, and resource anomalies through a complete monitoring alarm and processing system, automatically culling faulty disks for automatic processing, NIC faults, faulty nodes real-time sensing and shielding faults. Real-time push is performed through alarms for faults affecting system operation, and problems are discovered in advance. (Author: Jiang Wenlong)