Research of Self-adaptive Load-balancing Algorithm Based on Session

摘要

The self-adaptive load balancing algorithm based on session is introduced for the increasing load in the intrusion detection, then some proper aspects are selected to evaluate the algorithm performance. This algorithm concerns about allocating the new session to the engine which has the least loads by feedbacks. When load imbalance happens, data packets from the engine with heaviest loads are remapped to the one with the least loads. The results of our experiments in different environments show that this self-adaptive load balancing algorithm based on session is better than IBM self-adaptive algorithm when busy network traffic meets multi-engines.