Adaptive management method and device of digital monitoring platform memory

Abstract

The invention discloses an adaptive management method and device of a digital monitoring platform memory. The adaptive management method of the digital monitoring platform memory comprises the following steps: setting a memory pool, and configuring a minimum cache with a preset volume for each video channel; according to volume of a frame which needs to be written in at present, calculating the cache required for each video channel, and applying for the cache for each video channel according to the calculated cache; monitoring the code rate change of each video channel; when the code rate becomes big, continuing to apply for the cache; and when the code rate becomes small, releasing residual cache space to the memory pool. The use efficiency of the memory can be improved, and the storage performance of a NVR (Network Video Recorder) is improved.

Description

technical field [0001] The invention relates to the field of memory management, in particular to a memory adaptive management method and device for a digital monitoring platform. Background technique [0002] Video Storage Server (NVR), NVR is the abbreviation of (Network-Video-Recorder, Network Video Recorder). The main function of the video storage server is to receive the digital video stream transmitted by the IPC (network camera) equipment through the network, store and manage it, so as to realize the advantages of the distributed architecture brought by the network. Simply put, through the video storage server, multiple network cameras can be watched, browsed, played back, managed, and stored at the same time. [0003] With the development of high-definition video technology and information technology in the security industry, it is urgent to improve the performance of network video storage servers (NVR), and storage performance is one of the core performances of NVR ...

AI Technical Summary

Problems solved by technology

However, NVR uses a lot of memory, such as recording, pre-recording, playback, downloading, etc.

These memory usages are quite different. For example, the alarm pre-recording takes up to 20 seconds, and the normal pre-recording takes 3 seconds.

And even when there is only video recording, the bit rate of each front-end device changes quite a lot, for example, the bit rate may be as low as 256Kbps, and may be as high as 8Mbps

In many practical projects, server configuration, recording process, and alarm pre-recording may all change. If a fixed memory pool is used, it will cause different degrees of waste of server resources, or be set according to server hardware resources and different applications. This requires compiling different versions of NVR or modifying configuration files

As more and more NVRs are deployed on the platform, when NVR clusters appear, the previous methods are becoming more and more difficult to meet the requirements

Images