Research-on load balance in distributed network measurement system

With rapid development of network technology, network has penetrated into all fields of social life; it has evolved into a social infrastructure. Network measurement has been noted by network managers, and the analysis and store of mass measurement data make the necessity of distributed network measurement system. Load balancing among each node in the system becomes important. Given the strengths and weaknesses of traditional load balancing algorithm, we propose Weighed Probability Forwarding Scheduling to allocation requests of sending measurement data; it includes three aspects: scheduling granularity, weight assigning, and forwarding strategy. As is shown by experiments, it makes a good job of flow balance compared to traditional Round Robin Scheduling. And it also solves the mass requests problem.