With the rapid development of network and the emergence of new applications on IP network, new requirements are put forward for the quality of service of IP network. Real-time services such as VoIP(Voice over IP) put forward higher requirements for message transmission delay. If the transmission delay of the message is too long, users will not accept it (e-mail and FTP services are relatively insensitive to delay).
In order to support voice, video, data and other services with different business requirements, the network is required to distinguish different communications and provide corresponding services. In the traditional IP network, it is impossible to identify and distinguish various communication categories, and the ability to distinguish communication categories is the premise of providing different services for different communications, so the traditional network optimal service model can no longer meet the needs of applications. The emergence of QoS (Quality of Service) technology is dedicated to solving this problem.
QoS aims to provide different quality of service for different applications, such as providing dedicated bandwidth, reducing message loss rate, reducing message transmission delay and delay jitter. In order to achieve the above objectives, QoS provides the following functions: packet classification and coloring, network congestion management, network congestion avoidance, traffic monitoring and traffic shaping, and QoS signaling protocol.
The following two figures let us know the different processing processes of messages in the network without QoS guarantee and with QoS guarantee when the network is blocked.
Network without Qos application
All messages to be output by the interface enter the tail of the interface FIFO queue in the order of arrival. When the interface sends messages, it starts from the head of the FIFO (first in, first out) queue and sends messages in turn. There is no difference in the sending process of all messages, which does not guarantee the quality of message transmission.
Network supporting quality of service
After the message arrives at the interface, it is classified first, and then the message is allowed to enter the tail of the queue according to its category. When sending messages, according to the priority, the messages in the low priority queue are always sent after all the messages in the high priority queue are sent. In this way, every time a message is sent, the message with high priority is always sent out first, which ensures that the message belonging to the queue with high priority has low delay, and can also ensure two performance indexes of message loss rate and delay when the network is congested.
Maybe after reading the above introduction, you will have some questions. What does this have to do with consumers?
To sum up, if the QoS policy is not implemented, audio and video applications, such as IP telephony and video conferencing and mission-critical data, can only be transmitted as "best effort" services (non-guaranteed services), which will lead to the instability of voice and video when the network is congested. On the other hand, if the network is only used for a specific application system with no time limit, such as WEB or email, you don't need to select this function.