Technical solution implementation of high-definition surveillance storage

Abstract: As we all know, high-definition, as the overall trend of future video surveillance development, requires the cooperation and support of all links. For the storage of important links in the monitoring system, the application of high-definition monitoring has also put forward corresponding new requirements for its development: first, high-definition monitoring brings higher transmission bandwidth load and larger storage capacity to the system; second, " high-definition monitoring " often means a higher level of monitoring, and the reliability requirements for data storage are also higher. These new developments and changes have put forward new requirements for monitoring storage products and technologies.

Key factors for achieving high-definition storage

In order to better realize high-definition monitoring, storage technology needs to be matched in terms of capacity, performance, application response, structural adaptability and other aspects.

First, in the face of the wider data transmission bandwidth and larger storage capacity requirements of high-definition surveillance, storage products need larger capacity and higher performance. More importantly, when meeting these requirements, storage technology needs to be more closely integrated with surveillance applications, especially high-definition surveillance applications. For this, only products and solutions designed for specific applications can better meet the requirements.

Secondly, high-definition monitoring systems mostly use high-definition network cameras as video acquisition devices for monitoring points. Such devices often do not have long-term storage media themselves. The data storage process is more about transmitting data through the network and then storing the data on other specialized storage devices.

Third, while high-definition surveillance brings higher-definition image quality, it also brings with it the pressure of increased encoding streams and increased storage data volume with the same storage time. This requires storage devices to be able to efficiently and completely implement secure writing of video data based on the characteristics of high-definition surveillance code streams.

Finally, high-definition storage will change the previous characteristics of monitoring system data and encoding equipment mainly stored locally, and will centralize the data to one or more places for storage. In this way, how to quickly retrieve and obtain the required historical videos from the centralized massive data is also a new challenge.

In addition, systems that use high-definition encoding input generally use high-definition network cameras or ball cameras, and the system presents an easily expandable modular architecture. For this feature of the front end, the back-end storage solution also requires greater flexibility to meet the scalability requirements of the entire system.

Specific solutions for high-definition storage needs

With the increasing amount of HD surveillance data, higher requirements for storage reliability, and more flexible system structure, the problems that HD storage needs to solve are becoming more and more clear: in HD surveillance systems, storage devices and subsystems must have effective and complete recording functions; the reliability and security of the stored data must be considered from the perspective of the media and the system; the flexible system structure must be able to achieve unified management and scheduling of storage resources to configure flexible recording plans, and be able to respond to requirements for extended system recording time and expansion of the monitoring front end at any time according to changes in demand.

When overcoming these challenges, we focused less on the choice of storage media and more on designing different storage products based on existing online storage media disks, so as to meet the storage requirements of different high-definition surveillance systems in multiple dimensions:

To achieve large-capacity storage, products with multiple hard disk slots and compatibility with large-capacity hard disks can be designed, and massive storage space can be provided through the expansion of external array cabinets.

For high-reliability storage requirements, RAID protection technology can be used at the device level. For example, RAID 5 can support redundancy in the event of one damaged hard disk, and RAID 6 can support redundancy in the event of two damaged hard disks. The device uses a dual-controller architecture, using physical hardware redundancy and communication heartbeat association to ensure the reliability of storage services. At the system level, N+1 device online redundancy mode and dual-machine hot standby mode for video management servers can be used.

In terms of system architecture, it is recommended to adopt a distributed network storage solution to reduce the system transmission pressure in view of the characteristics of the monitoring system. Different storage subsystems should also be used for high-definition monitoring systems of different scales and structures. For example, for the transformation of old systems and partial implementation of high-definition systems, hybrid DVRs can be used to access and store high-definition videos, which not only protects the original analog standard-definition video access, but also adds a "high-definition" highlight to the system (as shown in Figure 1).

For the structure where the monitoring points of the system are relatively concentrated, a structure using iSCSI direct writing and direct access can be designed. Using embedded high-definition encoders and high-definition network front-ends (high-definition network cameras or high-definition network dome cameras) and IP-SAN, based on the iSCSI protocol, the monitoring video encoding end can be directly written to the storage end in units of data blocks. The storage resources of the entire monitoring system are uniformly managed by the storage management server to achieve unified scheduling and allocation of storage space required for encoding front-end recording. The storage space size can be flexibly divided and changed as needed, and storage resources of different sizes can be allocated to each encoding channel to meet the requirements of different monitoring points and different storage time lengths in the system. And it can achieve positioning accurate to the second level and the function of storing and releasing at the same time. When the monitoring client needs to play back or download operations, it can directly obtain the required data from the IP-SAN storage device based on authorization, and use the powerful data throughput capacity of IP-SAN to greatly improve the download performance and efficiency of recordings. The storage resources provided by the IP-SAN devices in the system are centrally managed by the storage management server, which can not only reduce the number of servers related to storage services, but also seamlessly switch the data storage area of ​​the monitoring point in the background when a single IP-SAN device fails, ensuring the continuous operation of the recording service (Figure 2).

In view of the complex network structure and hierarchical storage structure of the monitoring system, NVR devices can be deployed in different monitoring areas. Not only can the video collected by the high-definition monitoring points in the monitoring area be stored, but different recording types and time length strategies can be flexibly set for different monitoring points, providing data streaming video records, and based on the configuration of hot standby storage, when a single NVR fails, the backup NVR will take over its recording business (as shown in Figure 3).

The future of high-definition storage

As surveillance systems accelerate toward the high-definition era, storage, as an important part of system applications, is bound to have more development and changes.

Future surveillance storage will develop towards the trend of joint observation, response energy saving and low power consumption, just like the current general data storage. In addition to the energy-saving design of the equipment itself, there are more optimizations to be done in the storage process and system configuration. For example, adding storage MAID applications to keep non-read and write disks in a non-working state to reduce power consumption; by optimizing the recording writing method, the configuration of hardware devices in the system can be reduced to save power consumption.

In addition, in the future, video storage will no longer just store video data and related video index information, but also store additional information related to intelligent monitoring, because high-definition monitoring, in addition to bringing higher clarity, will also reflect more intelligent monitoring applications. How to achieve the association of video data and information data in the storage process to achieve intelligent applications and improve their application efficiency is of great practical significance.