Network flow packet capturing, flow limiting and packet loss method based on PID control algorithm

Abstract

The invention relates to the field of information technology processing, and discloses a network flow packet capturing, flow limiting and packet loss method based on a PID control algorithm, and the method comprises the steps: collecting three system load index quantities: a CPU utilization rate, an IO utilization rate and a memory utilization rate of a system where NTA applications are located; carrying out normalization, mean filtering and weighted fusion processing on the three index quantities to obtain a single characterization quantity; then taking the difference between the characterization quantity and a PID expected value as the input of a PID control system; finally, obtaining the packet loss control proportion, and discarding the overload data packets according to the proportion, so the stable operation of the system is maintained. Compared with a traditional fixed threshold control method, the method has higher flexibility, the load change of the system can be tracked in real time, the dynamic packet loss capability of different packet loss proportions according to different loads is realized, and higher service quality can be provided with less packet loss.

Description

technical field [0001] The invention relates to the field of information technology processing, in particular to a PID control algorithm-based method for packet capture, current limiting, and packet loss in a network. Background technique [0002] In the NTA system, in order to ensure the stable operation of the system, when the burst traffic arrives, it will consume excessive system hardware resources, resulting in overloaded operation. If the system works under overload for a long time, the probability of crash will increase. In order to ensure the smooth transition of the system when burst traffic occurs in the user network, the overloaded data packets are usually discarded. Dropping data packets is the most important technical response method in QOS. The most commonly used strategy in NTA's QOS is fixed-threshold overrun control. When the system hardware is determined, the upper limit of the total number of data packets that can be processed per time unit can be determ...

AI Technical Summary

Problems solved by technology

[0003] The overrun control method with a fixed threshold is simple and reliable, but there are also the following problems: Since different thresholds need to be maintained for different hardware, new hardware needs to test the threshold value, which increases the complexity of maintenance, and may be premature or If the action is too late, data packets that should not be discarded will be discarded or no action will be taken when the resources are exhausted. Furthermore, since the hard threshold does not examine the real-time hardware resource consumption and only uses the threshold as the only control indicator, so in real-time Insufficient responsiveness

Images