Data-oriented cloud storage two major trends
I. Overview
Looking at the current status of the security industry, the development of IP, high-definition, intelligent, and big data applications has been widely recognized with the landing of many landmark projects. SDT (security big data) will be the direction of future security development. Cloud storage technology as the basis for big data applications, while enhancing the storage system advantages of the security industry, the security market's demand for cloud storage is becoming more and more clear. At present, the scale of video surveillance systems continues to expand, as well as the large-scale application of high-definition video, the data that needs to be stored in video surveillance systems, and the complexity of data applications are constantly improving. Video data needs to be stored continuously in the storage system for a long time, and it is required to be called at any time. Therefore, strong demands are placed on the reliability and linear growth of the storage system.
As an industry with its own characteristics and characteristics, the security industry is full of expectation in the combination, adjustment and optimization of cloud storage technology and security industry characteristics. However, a general-purpose conventional cloud storage system separates cloud storage as a storage device and separates links with specific industries and specific applications. It only serves as a basic role for data read and write, storage, and security protection throughout the entire security system. By providing IT-based data read and write methods, the universal cloud storage system requires the application to adapt to the working mode of storage, thereby increasing the development tasks of the application, extending the business process of video storage, and inhibiting the reasonable optimization of cloud storage. Therefore, if cloud storage is to grow and develop in the security industry, in addition to inheriting its essential advantages, it should further understand the characteristics and needs of the security industry, and actively participate in the data path optimization work from the front end to the storage of the entire end-to-end data path. Come.
Second, the difference between cloud storage architecture
Currently, there are two types of cloud storage systems that are applied in the industry as a whole. There are two types of fully-symmetric cloud storage architectures and non-fully symmetrical cloud storage architectures. The hardware, software and functions of all nodes in the fully symmetrical cloud storage architecture are the same. That is, both the metadata module and the data module are in the same storage device. In this architecture, all storage devices have the same status. The cloud storage software system integrates the storage devices and provides services together. The non-fully symmetrical cloud storage architecture mostly uses the metadata module deployed on the server independently. The data module is deployed on the storage device. The cloud storage software system controls the overall service of the system and separates the signaling flow from the service flow. Cloud storage service. Although there are major differences in architecture, the nature of cloud storage provisioning services is not much different. Unified service provision capabilities, virtualized space uغير مجاز مي باشدe, distributed storage models, real-time dynamic business scheduling, and disaster recovery are several aspects of measuring the basic attributes of cloud storage.
Third, security cloud storage development direction
With the above characteristics, although the storage system itself has completed the process of cloudization, it has further differentiated itself after entering the security industry, and has evolved into two different paths when combined with the characteristics of the security business.
1. "Used forism"
One road was taken to bring the universal cloud storage system on the market into the security industry. The original intention of this method is to use cloud storage systems to facilitate management, simple use, high scalability, and high reliability for security services. However, when it is actually used, it is not so simple. Take video recording as an example. When traditionally using an NVR as a storage device, the management platform only needs to issue video commands to send the video stream directly to the NVR device for storage. Between the camera and the NVR Without additional data transfer equipment, the process is simple and the storage failure points are only concentrated in the NVR itself. Although the introduction of general-purpose cloud storage can improve the reliability of the storage side, it can no longer meet this short-path data storage. The reason is that a general-purpose cloud storage system does not have the ability to contact the camera to complete its own stream.
The general-purpose cloud storage system provides services by connecting its own capacity space to a device through a protocol or API interface. Therefore, to complete the entire process of video recording, monitoring industry platforms need to introduce independent devices to build cameras and cloud storage. The bridge between. The common practice is to increase the streaming media server and the store and forward server. The streaming media is used for streaming forwarding and the store and forward server is used for interfacing with the universal cloud storage. Therefore, the whole process of data flow consists of: camera → NVR, camera → streaming server → store and forward server → cloud storage, and the entire data path is elongated. Although cloud storage has replaced NVRs in terms of management, performance, expansion, security, and other aspects, the protection of newly introduced links cannot be discussed, and security vendors are required to protect them.
In addition, data written in general-purpose cloud storage is mostly stored in files. The server converts a continuous stream of continuous video streams into a large number of interrupted video clip files and stores them in a cloud storage system. The characteristic of video data is streaming data. Generating files will have certain influence on the flow characteristics of data and reduce the efficiency of data stream writing. At the same time, when using and searching video files, most of them use file names to search, and positioning is not Very accurate. Therefore, although the application of the general-purpose cloud storage system in the security industry has improved the storage system, the newly introduced problems also require security vendors to overcome it.
2, "Student's skills to strengthen themselves"
The other road of cloud storage entering the security industry can be summarized as "Leading yourself to take advantage of longevity," which inherits the essential advantages of cloud storage, and at the same time, it is adapted to the security industry to adapt to the actual situation of the security industry.
(1) Overall system architecture adjustment In adapting to the actual situation in the security industry, first, the cloud storage system can independently assume the functions of fetching, previewing, and forwarding of video images through its own transformation. In this way, in the entire system, the data flow can be directly written from the camera to the storage node of the cloud storage system, and the distributed storage of data is completed by the storage nodes in the cloud storage system. The entire structure eliminates a lot of intermediate links. It only involves the front-end and storage. As a result, the number of links is reduced, the efficiency of the flow is simplified, and the failure links are reduced.
In addition, the cloud storage system can also perform application-oriented customization for video storage technologies. For example, periodic management of video data, video recording, index management, and video integrity detection can be combined with various storage-related content into cloud storage. system. This frees the system resources of the upper business system and simplifies the entire security storage structure.
(2) Optimization of Video Recording Efficiency The cloud storage system can also optimize the data characteristics of the video. In order to avoid file-based storage methods, video data is artificially cut into a large number of video clips to destroy its integrity, and the storage method of video data can also use streaming storage in cloud storage. This of course requires the cloud storage system to understand in depth the characteristics of storage and use of video data in the security environment, and to transform its own storage methods.
Video recording data for 30 days, for example, the original source of continuous video data will form a 30-day data volume for the total capacity, time circumference. The data is repeated over the 30-day cycle and is continuously written. Therefore, for the data characteristics of video recording, the cloud storage system can no longer use non-streaming data such as video clip files, and instead uses a storage method that is more in line with the characteristics of streaming data. A camera's data for 30 days is stored in the form of a circle, the data volume is the total circumference of the full circle, and the time coordinates are marked on the circumference in units of seconds or finer granularity. So, without destroying the data integrity, you can customize the data positioning, query, download, and playback according to the precise coordinates of the time. The storage and use of video data will be greatly improved.
In the security industry, one of the advantages of cloud storage is reflected in the big data platform, as well as its own transformation and optimization based on security application characteristics. For example, Hikvision's video cloud storage is characterized by long-term non-interrupted generation of video data and periodic coverage. It also creatively develops a streaming data structure and stores video data in a more efficient and relevant manner, making data storage more efficient and data-rich. The retrieval is faster and the long-term use performance of the system does not decrease. Cloud storage performs preliminary analysis of the key information of the data and stores it together with the data. This will no longer be restricted when the data is used for intelligent analysis and application expansion, which is more conducive to the development of security big data. Therefore, the security innovation that comes with it will be more flexible and richer. Hikvision's video cloud storage and Hikvision Smart's front-end and platform collaboration can provide such features as: regional intrusion detection, picture comparison retrieval, and video enrichment. And other rich security applications.
Fourth, summary
For security vendors, the cloud storage system is introduced as a task to improve the storage end. Therefore, more practical considerations must be devoted to the practical application of cloud storage in the security industry. No matter what kind of cloud storage mode, what kind of technical route can be used to complete basic storage tasks, but there are big differences in how to improve overall efficiency and reduce extra consumption while completing data storage tasks. Therefore, while leveraging the unique advantages of cloud storage, focusing on the specific characteristics of specific industrial data, focusing on the development needs of future applications, and continuously integrating and optimizing, it is necessary to form a more secure cloud storage system that fits the needs of users and is more vital. system.
Cold Drawn Seamless Steel Pipe
Steel Tubing,Stainless Pipe Co., Ltd. , http://www.czsteelpipes.com
برچسب: ،
ادامه مطلب
