Research on key technologies of technological service and management based on cluster load balancing

摘要

AbstractDetection of similarity in scientific and technical data finds repetition rate of contents in project notification form for technology plan submitted by each unit. It is conducted through comparison between the same batch and previous batches on declaration book from each declaration unit. Furthermore, it affords early warning, which is an important link in normativeness for project management. Usually, cluster system based on Web servers is adopted to reduce data processing time, where loading balancing of each node is a research hotspot. This paper proposed auto-adaptive load balancing (AALB) algorithm to accomplish load balancing for each node in the cluster by predicting request arrival rate and request size and rapidly adjusting parameter value. The experiment proves that data block distribution is more balanced in each node using balancing strategy for AALB, with less time compared with other load balancing.