In high-definition network video surveillance systems, there are often customer feedback screen delays, catastrophic phenomena, etc. There are many reasons for this phenomenon, but in most cases, the configuration of the switch is not reasonable enough, resulting in insufficient bandwidth. In terms of network topology, a medium and large HD network video surveillance system needs to adopt a three-layer network architecture: access layer, aggregation layer, and core layer.
1. Selection of access layer switches
The access layer switch mainly connects the front-end network HD camera and the uplink aggregation switch. Calculated by 4M code stream of 720P network camera, can a 100M port access switch access a few 720P network cameras?
The actual bandwidth of our commonly used switches is 50%-70% of the theoretical value, so the actual bandwidth of a 100M port is 50M-70M. 4M*12=48M, so it is recommended that one 100M access switch can access up to 12 720P network cameras.
At the same time, considering the current network monitoring adopts dynamic coding mode, the peak value of the camera code stream may exceed 4M bandwidth, and the bandwidth redundancy design is considered. Therefore, it is best to control one 100M access switch within 8 units, and more than 8 recommended. Gigabit.
2. The choice of aggregation layer switch
The aggregation layer switch is mainly connected to the access layer switch and uplink control center core switch. In general, an aggregation switch needs to select a Layer 2 switch with a Gigabit uplink.
It is still calculated by the 4M code stream of the 720P network camera. There are 6 720P network cameras on each access layer switch in the front end, and the aggregation switch is connected to 5 access layer switches. The total bandwidth of the aggregation switch is 4M*6*5=120M. Therefore, the aggregation switch and the core switch should be connected to the Gigabit interface.
3.The choice of the core layer switch
The core layer switch is mainly connected to the aggregation layer switch, the uplink monitoring center video monitoring platform, the storage server, the digital matrix and other devices, and is the core of the entire high-definition network monitoring system. In selecting the core switch, it is necessary to consider the bandwidth capacity of the entire system and how the core layer switch is improperly configured, which inevitably results in a video display that cannot be smoothly displayed. Therefore, the monitoring center needs to select a full Gigabit core switch. If there are more points, you need to divide the VLAN. You should also choose a three-layer full Gigabit core switch.
Attachment: Parameters that determine the performance of the switch
Backplane bandwidth calculation method: number of ports * port speed * 2 = bandwidth of the backplane. Take Huawei S2700-26TP-SI as an example. The switch has 24 100 Mbps ports and two Gigabit uplink ports. Backplane bandwidth = 24*100*2/1000+2*1000*2/1000=8.8Gbps.
Packet forwarding rate:
Packet forwarding rate calculation method: full configuration GE port number × 1.488 Mpps + full configuration 100 Mbps port number × 0.1488 Mpps = packet forwarding rate (1 gigabit port when the packet length is 64 bytes, the theoretical throughput is 1.488 Mpps, The theoretical throughput of a 100 Mbps port at a packet length of 64 bytes is 0.1488 Mpps). Take Huawei S2700-26TP-SI as an example. The switch has 24 100 Mbps ports and two Gigabit uplink ports.
Packet forwarding rate = 24 * 0.1488 Mpps + 2 * 1.488 Mpps = 6.5472 Mpps.