A Grid Accounting Model Based on Event-Trigged Mechanism and User Behavior

Accounting system is one of the building blocks in the grid computing and cloud computing environments. Despite its growing significance, the quality of accounting service suffers from several problems, including the tight coupling relationship between the accounting methods and the accounting log files, slow updates, and ignorance on the impact of user behavior. To address the above problem, an accounting model based on event-trigged mechanism and user behavior was designed. The event-trigged mechanism was used to collect the accounting information. This method relieves the tight coupling relationship between the accounting methods and the log files, thus improves the speed and efficiency of collecting the accounting information. The model also introduces an accounting policy based on users´ behavior. In this policy, the user behavior is evaluated, and the user´s weight factor is computed, consequently, the user could get reward or punishment accordingly. The experimental results show that the proposed accounting model reduces the average time delay of collecting the accounting information, differentiates incentives according to user behavior, and improves the quality of accounting services.