К вопросу информационной безопасности Grid-систем

Abstract

We review different approaches to enabling security in Grid systems. We analyze such mechanisms as Globus GSI and message-level security. We also focus on the following aspects of Grid security: anomaly identification, and resource and data availability monitoring. Proposed model of user behaviour for anomaly identification is based on the analysis of statistical data about jobs that were carried out by the user in Grid system. In order to distinguish normal and abnormal user’s behaviour we use feed-forward neural networks. We verified the proposed approach on data that were collected by GridICE monitoring system in GILDA-EGEE. The results of experiments showed that our model was able to distinguish abnormal user’s behaviour in 90%. Resource availability monitoring is based on the concept of active experiment. This enables the analysis of availability of Grid resource in real-time