At present, most manufacturers have introduced IP video surveillance systems using this model. The core of this model is to use the independent streaming media server in the system or the streaming media function module in a device to realize the video stream replication distribution, so as to achieve video client decoding and playback, video decoding on the wall, and storage in the system The server or the storage function module obtains the video forwarded by the streaming media server to implement video storage. This model itself has also undergone a series of evolution and development.
The storage server and streaming media server at this time are all high performance computers. The streaming server obtains the video stream from the front-end camera, then copies the video stream and distributes it to the storage server. Since the storage requirements in the IP surveillance system are basically real-time storage throughout the day, the video stream distributed to the video storage server will continue to exist. If the client software or decoder requires real-time video streaming on the wall, the streaming media server will copy one or more video streams to the client and the decoder on the wall.
The streaming server obtains the video stream from the front-end camera, and then copies the video stream, which will surely be distributed to the storage server. Since the storage requirements in the IP surveillance system are basically real-time storage throughout the day, the video stream distributed to the video storage server will continue to exist. If the client software or decoder requires real-time video streaming on the wall, the streaming media server will copy one or more video streams to the client and the decoder on the wall.
In this structure, the work pressure is mainly on the streaming media server. The forwarding capacity of a server is limited. If the system is a high-definition camera, the number of forwarding will be significantly reduced. Let's say storage, the storage function of the system is mainly completed by the storage server and the disk array. The role of the storage server is to obtain the video stream from the streaming media server, then package it into a file format and send it to the disk array for storage. Here the storage server and the disk array There will be two connections: one is through the IDE or SATA cable directly connected, that is, DAS; the other is through the network, that is, NAS / IPSAN.
The biggest problem with the above structure is that the number of servers in the system will be large. This is especially true for large-scale monitoring systems with multiple points. This obviously increases the cost and maintenance complexity of the system. At the same time, because the streaming media server and the storage server are ordinary PC-style servers, the running program is basically based on WINDOWS development, and it also has certain hidden troubles in its stability.
Integrated structure of streaming media module and storage module
The improved structure of the IP video surveillance system is mainly to integrate the streaming media server and the storage server as two independent functional modules on one server. This will not only reduce the number of servers in the system, but also increase the number of servers in the system. The internal bus transfers the video stream to the storage module to reduce the network bandwidth pressure. At the same time, the storage module obtains the video stream forwarded by the streaming media module is more reliable and stable. However, after the storage module processes the video data into file packages, it will still be transmitted to the disk array storage through the network, which still consumes network bandwidth resources.
Join the structure of embedded NVR
In order to improve the stability of the storage part, embedded NVR appeared. The embedded NVR integrates the original NVR server and disk array in a structure, which is generally composed of a server head plus several disk storages. The software in the system is also changed from the previous WINDOWS-based storage software to an embedded software. More stable and reliable, with the advent of embedded NVR, a considerable part of the structure of the IP surveillance system evolved into the form described in Figure 3.
Since the early embedded NVR only had the storage function and did not have the function of forwarding video, the streaming media server in the system continued to exist, but the storage part became an integrated embedded NVR device, in addition to the storage operation is more stable and reliable, NVR After the video stream from the streaming media is obtained, the remaining work is completed in the local computer. The video data is no longer sent to the network and transferred to the independent disk array. This reduces the network bandwidth pressure.
Structure without streaming media forwarding server
The embedded NVR soon became a very important part of the IP surveillance system. In addition to the storage functions, more features were added to the embedded NVR. The most important of these was the video streaming and video management functions. The original system The streaming media forwarding server will no longer be needed. The video management function enables the embedded NVR to have the ability to form a small system independently. In a project like a residential area or a chain store, the embedded NVR is the core of the system and it has all the IP digital monitoring system. The main function, in large systems, embedded NVR will be integrated into the entire system as a basic component. This is also one of the mainstream IP monitoring system structures.
In addition to the management server is indispensable in the system, the embedded NVR becomes the basic unit of the system, which has video forwarding and storage capabilities. These NVR units are configured to directly obtain video streams from the front-end IP cameras under their jurisdiction. If the outside world does not have the need for real-time browsing, these NVRs directly turn these video streams into file packages and store them in the disk array in the local machine. Or the decoder's real-time browsing needs, in response to these requirements, copy another or several video streams forwarded to the client software or decoder. The structure of the entire system is simpler and clearer, and the bandwidth pressure of the network is also greatly reduced.
The above structures are essentially the same in nature and are based on streaming media forwarding technology to achieve browsing and storage. There are two problems with these types of structures:
The browsing video stream and the storage video stream originate from the same source and are not flexible enough to be applied. Specifically, in the streaming media forwarding architecture, the streaming media section (whether it is a function module or a stand-alone device) only acquires one from the front end. The video stream is then forwarded to the storage or browsing device. If the front-end camera is a high-definition camera, and the user saves high-definition video, then the browsing must also be high-definition video. A client computer decoding more than 9 HD video may not be enough. In addition, if the customer's storage space is limited and he wants to browse high-definition video but store standard definition video, it would be very difficult to achieve this structure without special treatment. A more practical requirement is that high-definition video needs to be stored, but browsing does not need to always be high-definition video. When a client opens 9 or 16 frames, whether a single frame is HD can not be resolved. Display SD or smaller resolution video, the client computer will be more relaxed when decoding these non-HD video, the fluency of the picture is also higher, when you switch back to a single screen, you only need to display high-definition video.
Currently there are two main ways to solve this problem.
First, the streaming media part detects the number of client-side multi-screens through the management server. Once the client is found to be set to more than 9 screens, the streaming media module will cut the HD video stream to a low-resolution video forwarding client. When the client resumes the single-image window, the high-resolution video stream is resent. But doing so will further increase the burden on the streaming media module. In the case of a certain total resources, it will inevitably affect the ability to copy and forward video streams. At the same time, the high-definition video stream of the front-end camera is also preferably multi-level and can be tailored.
Another method is to use another stream from the front camera. At present, high-definition cameras generally support at least one high-definition stream and one low-resolution stream output. When the streaming media module detects that the client opens a multi-picture window, Re-acquire a low-resolution video stream from the front-end camera for forwarding, and disconnect the originally forwarded HD video stream. This sometimes causes short-term video-free phenomenon when the client performs multi-screen single-screen switching. This may be more noticeable when the device is transmitting video.
NVR storage mode is not flexible enough
In this structure, each NVR will manage a certain amount of front-end video, specifically, each number of video is stored in an NVR device. Although the embedded NVR is much more stable than the previous PC-based NVR, if a certain NVR fails, several front-end videos managed by this NVR cannot be recorded. Then the N+1 mode is used to make this problem possible. A certain degree of resolution. The N+1 mode is that in addition to the necessary number of NVRs, one or more (usually one) NVRs are prepared in the system. Usually, this NVR does not work but is in a ready state. Once the management server detects a certain number of NVRs. When one NVR fails or goes offline, it sends an instruction to the hot-backup NVR. The hot-backup NVR takes the initiative to take over the affected front-end camera and saves the video data in the hot-backed NVR. At the same time, the system warns the maintenance personnel to check and repair. Faulty equipment. Once the original faulty NVR is repaired or re-initiated, the hot-backed NVR will send the video stored in the unit back to the original NVR through the network. At the same time, the original NVR will re-take over the relevant camera. After the hot-spare NVR finishes sending the video data, Continue to be hot standby. But what if there are more NVR failures in the system? In the end, how many hot NVRs? Currently, mainstream vendors only support the N+1 mode, which means that only one NVR is allowed to fail.
The storage server and streaming media server at this time are all high performance computers. The streaming server obtains the video stream from the front-end camera, then copies the video stream and distributes it to the storage server. Since the storage requirements in the IP surveillance system are basically real-time storage throughout the day, the video stream distributed to the video storage server will continue to exist. If the client software or decoder requires real-time video streaming on the wall, the streaming media server will copy one or more video streams to the client and the decoder on the wall.
The streaming server obtains the video stream from the front-end camera, and then copies the video stream, which will surely be distributed to the storage server. Since the storage requirements in the IP surveillance system are basically real-time storage throughout the day, the video stream distributed to the video storage server will continue to exist. If the client software or decoder requires real-time video streaming on the wall, the streaming media server will copy one or more video streams to the client and the decoder on the wall.
In this structure, the work pressure is mainly on the streaming media server. The forwarding capacity of a server is limited. If the system is a high-definition camera, the number of forwarding will be significantly reduced. Let's say storage, the storage function of the system is mainly completed by the storage server and the disk array. The role of the storage server is to obtain the video stream from the streaming media server, then package it into a file format and send it to the disk array for storage. Here the storage server and the disk array There will be two connections: one is through the IDE or SATA cable directly connected, that is, DAS; the other is through the network, that is, NAS / IPSAN.
The biggest problem with the above structure is that the number of servers in the system will be large. This is especially true for large-scale monitoring systems with multiple points. This obviously increases the cost and maintenance complexity of the system. At the same time, because the streaming media server and the storage server are ordinary PC-style servers, the running program is basically based on WINDOWS development, and it also has certain hidden troubles in its stability.
Integrated structure of streaming media module and storage module
The improved structure of the IP video surveillance system is mainly to integrate the streaming media server and the storage server as two independent functional modules on one server. This will not only reduce the number of servers in the system, but also increase the number of servers in the system. The internal bus transfers the video stream to the storage module to reduce the network bandwidth pressure. At the same time, the storage module obtains the video stream forwarded by the streaming media module is more reliable and stable. However, after the storage module processes the video data into file packages, it will still be transmitted to the disk array storage through the network, which still consumes network bandwidth resources.
Join the structure of embedded NVR
In order to improve the stability of the storage part, embedded NVR appeared. The embedded NVR integrates the original NVR server and disk array in a structure, which is generally composed of a server head plus several disk storages. The software in the system is also changed from the previous WINDOWS-based storage software to an embedded software. More stable and reliable, with the advent of embedded NVR, a considerable part of the structure of the IP surveillance system evolved into the form described in Figure 3.
Since the early embedded NVR only had the storage function and did not have the function of forwarding video, the streaming media server in the system continued to exist, but the storage part became an integrated embedded NVR device, in addition to the storage operation is more stable and reliable, NVR After the video stream from the streaming media is obtained, the remaining work is completed in the local computer. The video data is no longer sent to the network and transferred to the independent disk array. This reduces the network bandwidth pressure.
Structure without streaming media forwarding server
The embedded NVR soon became a very important part of the IP surveillance system. In addition to the storage functions, more features were added to the embedded NVR. The most important of these was the video streaming and video management functions. The original system The streaming media forwarding server will no longer be needed. The video management function enables the embedded NVR to have the ability to form a small system independently. In a project like a residential area or a chain store, the embedded NVR is the core of the system and it has all the IP digital monitoring system. The main function, in large systems, embedded NVR will be integrated into the entire system as a basic component. This is also one of the mainstream IP monitoring system structures.
In addition to the management server is indispensable in the system, the embedded NVR becomes the basic unit of the system, which has video forwarding and storage capabilities. These NVR units are configured to directly obtain video streams from the front-end IP cameras under their jurisdiction. If the outside world does not have the need for real-time browsing, these NVRs directly turn these video streams into file packages and store them in the disk array in the local machine. Or the decoder's real-time browsing needs, in response to these requirements, copy another or several video streams forwarded to the client software or decoder. The structure of the entire system is simpler and clearer, and the bandwidth pressure of the network is also greatly reduced.
The above structures are essentially the same in nature and are based on streaming media forwarding technology to achieve browsing and storage. There are two problems with these types of structures:
The browsing video stream and the storage video stream originate from the same source and are not flexible enough to be applied. Specifically, in the streaming media forwarding architecture, the streaming media section (whether it is a function module or a stand-alone device) only acquires one from the front end. The video stream is then forwarded to the storage or browsing device. If the front-end camera is a high-definition camera, and the user saves high-definition video, then the browsing must also be high-definition video. A client computer decoding more than 9 HD video may not be enough. In addition, if the customer's storage space is limited and he wants to browse high-definition video but store standard definition video, it would be very difficult to achieve this structure without special treatment. A more practical requirement is that high-definition video needs to be stored, but browsing does not need to always be high-definition video. When a client opens 9 or 16 frames, whether a single frame is HD can not be resolved. Display SD or smaller resolution video, the client computer will be more relaxed when decoding these non-HD video, the fluency of the picture is also higher, when you switch back to a single screen, you only need to display high-definition video.
Currently there are two main ways to solve this problem.
First, the streaming media part detects the number of client-side multi-screens through the management server. Once the client is found to be set to more than 9 screens, the streaming media module will cut the HD video stream to a low-resolution video forwarding client. When the client resumes the single-image window, the high-resolution video stream is resent. But doing so will further increase the burden on the streaming media module. In the case of a certain total resources, it will inevitably affect the ability to copy and forward video streams. At the same time, the high-definition video stream of the front-end camera is also preferably multi-level and can be tailored.
Another method is to use another stream from the front camera. At present, high-definition cameras generally support at least one high-definition stream and one low-resolution stream output. When the streaming media module detects that the client opens a multi-picture window, Re-acquire a low-resolution video stream from the front-end camera for forwarding, and disconnect the originally forwarded HD video stream. This sometimes causes short-term video-free phenomenon when the client performs multi-screen single-screen switching. This may be more noticeable when the device is transmitting video.
NVR storage mode is not flexible enough
In this structure, each NVR will manage a certain amount of front-end video, specifically, each number of video is stored in an NVR device. Although the embedded NVR is much more stable than the previous PC-based NVR, if a certain NVR fails, several front-end videos managed by this NVR cannot be recorded. Then the N+1 mode is used to make this problem possible. A certain degree of resolution. The N+1 mode is that in addition to the necessary number of NVRs, one or more (usually one) NVRs are prepared in the system. Usually, this NVR does not work but is in a ready state. Once the management server detects a certain number of NVRs. When one NVR fails or goes offline, it sends an instruction to the hot-backup NVR. The hot-backup NVR takes the initiative to take over the affected front-end camera and saves the video data in the hot-backed NVR. At the same time, the system warns the maintenance personnel to check and repair. Faulty equipment. Once the original faulty NVR is repaired or re-initiated, the hot-backed NVR will send the video stored in the unit back to the original NVR through the network. At the same time, the original NVR will re-take over the relevant camera. After the hot-spare NVR finishes sending the video data, Continue to be hot standby. But what if there are more NVR failures in the system? In the end, how many hot NVRs? Currently, mainstream vendors only support the N+1 mode, which means that only one NVR is allowed to fail.
Rainfall Shower Head, Hand Held Shower Head, Stainless Steel Shower Head, Bathroom Shower Head
ZHEJIANG KINGSIR VALVE CO., LTD. , https://www.kingsirvalves.com