Large-scale video surveillance systems will be dominated by cloud storage

 

With the popularization of HD technology, 720 p and P videos have blossomed everywhere. Meanwhile, based on the pursuit of clarity, smart analysis and processing, cameras with 500 million, 800 million, or even tens of millions of higher resolutions are emerging. The problems with such high-definition monitoring are also emerging. HD video data may easily contain several GB to dozens of GB of files, higher requirements are imposed on the capacity, read/write performance, reliability, and scalability of storage devices, and they are also facing greater challenges for storage vendors; it is especially important to choose an appropriate data storage solution in the video monitoring system. From the perspective of system construction, we need to fully consider functional integration, data security, data stability, system scalability, performance, and cost. The high-definition monitoring system has high bit rate, large data volume, and high reliability requirements, mainly reflected in the following aspects:
High concurrent read/write capability, retrieval and Intelligent Analysis of massive video data, reliable and secure data protection capability, and flexible and convenient space expansion capability. In the face of the challenges posed by the above high-definition storage, let's take a look at the storage solutions:
　I. node-type Storage System
The node nvr architecture is mainly designed for small HD monitoring systems. The number of HD front-ends is generally less than dozens of channels. There is no large data center for storage and monitoring in system construction. The storage capacity is relatively small, and the requirements for user experience and system function integration are high.
To cope with this monitoring system, tiandiweiye M7 series node-type nvr integrates dual-HDMI local HD display, proxy forwarding, rich UI interface and smart analysis functions to not only store 16 channels of HD video images, furthermore, we have made in-depth Optimizations in Intelligent Analysis of video data and quick search and positioning. This architecture is designed for small-sized systems. There are few channels to access the HD front-end, and video data cannot be shared between different nodes. However, the system features are highly integrated and can be deployed as distributed nodes in large systems.
At the market application level, supermarkets, shops, small enterprises, and basic management units in the Political and Legal industries are widely used.
　　Ii. centralized storage system
The Platform SAN architecture is mainly designed for medium and large HD monitoring systems, with hundreds or even tens of thousands of front-end routes. Generally, IP San or FC San is used to build a high-definition video storage system. As an important part of the monitoring platform, front-end monitoring data is stored in San through the video storage management module.
This architecture improves the number of high-definition front-end channels relative to the node nvr, and provides fast and convenient scalability and mature technology. For IP San, although the data concurrent read/write transmission rate in the iSCSI stage is consumed, it has advantages such as good scalability, general hardware platform, and full sharing of massive data, it is still favored by many customers. Tiandiweiye IP San hardware is equipped with redundant power supplies and supports raid0, 1, 5, and 6 technologies to ensure the reliability and security of HD video data. Multiple network ports support the load balancing function, it can effectively cope with concurrent read/write of HD systems.
FC San is widely used in industry users and closed storage systems, such as county-level or prefecture-level city HD monitoring projects. Concurrent reading and writing of large data volumes poses great challenges to gigabit network switching, however, building a relatively independent storage subsystem using FC San can effectively solve the above problems.
In the face of the features of the video monitoring system with large files and random reads and writes, the data sharing redundancy between different storage units in the platform SAN architecture system remains to be improved; from the high-performance server's video data forwarding to the storage space policy, the system architecture also increases the hidden fault points and iSCSI bandwidth bottlenecks, resulting in insufficient utilization of hardware data concurrency performance and fewer front-end data access. The above issues gave rise to the platform nvr architecture solution.
The Platform nvr architecture uses a storage system as the core link for data management and storage. This reduces the deployment of storage servers, and the front-end video monitoring data is directly written into the storage unit, the monitoring platform and client can directly read or call video data from the storage unit.
This solution saves storage servers in the system architecture and eliminates the performance bottleneck and spof risks mentioned above. It greatly improves the writing and retrieval speed of the storage system, and completely eliminates problems such as file system damage caused by power supply and network instability of traditional file systems.
The data stored in the platform nvr can be queried by multiple clients at any time, on-demand. When you need to view multiple saved video monitoring data, you can use an authorized video monitoring client to directly query and display the video monitoring data at the corresponding location for viewing historical images. Because the data management server has the index of the video files of all monitoring sites in the monitoring system, the video monitoring client can query and VOD the data of all monitoring sites in the monitoring system through the CMS authorization on the platform, this process is transparent to users.
With the widespread popularity of HD systems, platform-based nvr gradually occupies a more important position. Suitable for medium and large monitoring systems in terms of deployment;
　　Iii. cloud-based storage solutions
Cloud storage is a new concept that extends and develops in the cloud computing concept. It refers to the use of cluster applications, grid technology, distributed file systems, and other functions, A system that integrates a large number of different types of storage devices in the network through application software to jointly provide data storage and business access functions. With the development of network technology and the improvement of bandwidth, the prospect of cloud storage in video monitoring is becoming wider and wider. Cloud storage, as a benchmark for future development, has the following advantages:
1. cost advantages: cloud storage is not a storage, but a service. Just like cloud-based Wan and the Internet, cloud storage does not refer to a specific device for users, it refers to a collection of many storage devices and servers. A user uses cloud storage instead of a storage device, but a data access service provided by the entire cloud storage system. In this way, users do not need to invest more in backend storage devices such as servers and hard disks. They only need to invest part of the service fee, greatly reducing the cost;
2. Intelligent storage and higher efficiency: All storage resources are integrated and users can see a single storage space. Meanwhile, cloud storage greatly improves storage efficiency, the virtualization technology solves the waste of storage space and can automatically re-allocate data to improve the utilization of storage space. For example, if a user needs a TB storage space, he only needs to rent a TB service, to increase space, you only need to increase the rental space;
3. Data retrieval is faster. The system generates Pb-level massive monitoring data, and the number of storage devices reaches dozens and hundreds. Therefore, scientific and efficient management is very important. Cloud storage provides centralized management tools for multiple devices based on the cluster management technology. It provides the functions of centralized device monitoring, cluster management, system software and hardware running status monitoring, active alarms, and image system detection. In the application of massive video storage and retrieval, the retrieval performance is particularly important. In traditional file systems, file retrieval uses the "directory-> sub-directory-> file-> location" search procedure. In the high-definition video monitoring of massive data, the number of directories and files is very large, and the efficiency of this search mode will be greatly reduced. This problem can be effectively solved by locating the serial number file.
4. Cloud storage provides high system redundancy and security. When the online storage system fails, the hot standby machine can take over the service immediately. When the fault is restored, the service and data are migrated back. If the faulty machine data needs to be called, you can insert the disk of the faulty machine into the cold backup machine to enable immediate availability of all data.
As a trend in future security storage development, cloud storage vendors are currently combining various search, application technologies, and cloud storage technologies to provide enterprises with a series of data services. However, the development of cloud storage is still subject to the following factors:
1. With network bandwidth restrictions, a real cloud storage system will be a large public service system distributed across multiple regions, all over the country, and even all over the world, users need to connect to cloud storage through broadband access devices such as ADSL and DDN. Only when the broadband network is well developed can users obtain sufficient data transmission bandwidth to transmit large amounts of data and truly enjoy cloud storage services. Otherwise, they can only talk about it;
2. Data Security: Since cloud computing was born, security has always been one of the top priorities for enterprises to implement cloud computing. Similarly, in terms of cloud storage, security is still considered on the homepage, for customers who want to perform cloud storage, security is usually the top commercial and technical considerations. However, many users have higher security requirements for cloud storage than their own architecture can provide.
In this case, many large and reliable cloud storage vendors are trying to meet their requirements in the face of such unrealistic security requirements, build a data center that is much safer than most corporate data centers and protect data in storage through an encryption layer and protection layer comparable to the NSA. Today, users often find that cloud storage has fewer security vulnerabilities, and the security level provided by cloud storage is higher than that provided by their own data centers.
3. Application storage development: cloud storage is more than just storage, but also applications. Application storage is a storage device that integrates application software functions in storage devices. It not only has data storage functions, but also has application software functions. It can be seen as a collection of servers and storage devices. The development of application storage technology can greatly reduce the number of servers in cloud storage, thus reducing system construction costs, reducing single point of failure and performance bottlenecks caused by servers in the system, and reducing data transmission links, provides system performance and efficiency to ensure efficient and stable operation of the entire system.
At the same time, cloud storage must face the limitations of Web2.0 technology, application storage technology, file system, P2P, data compression, CDN technology, and virtualization technology. Therefore, cloud storage technology is mature, it takes a long time to go;
　　Conclusion
With the development of HD monitoring, users' requirements are clearly visible; massive HD video data has high requirements on the stability, scalability, retrieval capability, energy conservation, environmental protection, and cost of the storage system. There will be two trends in the future security market: one is a small-scale video monitoring system, mainly nvr; the other is a large-scale video monitoring system, which will be dominated by cloud storage.

Large-scale video surveillance systems will be dominated by cloud storage