Abnormal Access Request Detection based Secure Storage Area Network System

In order to ensure security of the enormous data stored in the storage area network system, current secure technology need large time and space consumption and extra storage space. The artificial immune algorithm can efficiently detect abnormal data. In order to keep high I/O performance of storage area network system, we use the artificial immune algorithm to detect the abnormal access request of the storage area network system, and we implement the secure storage area network system named ADSSAN. We present the structure of the metadata server with the central abnormal access request detection module, the storage device with the associate abnormal access request detection module and the secure storage area network system. The definitions of main elements are given. By analyzing the feature of the storage area network system, the master-slave artificial immune algorithm is presented. The central abnormal access request detection module generates all detectors and distributes part of detectors to the associate abnormal access request detection module in the storage servers. Then the distributed strategy of artificial immune algorithm can be used to reduce the time and space consumption of generating detectors and improve performance of abnormal access request detection. The prototype of abnormal access request detection module is implemented and tested, the result shows that the master-slave artificial immune algorithm can accurately detect abnormal access request with low time and space overhead. At last, the prototype of secure storage area network system is implemented on Lustre and evaluated its I/O performance. The result shows that the central and associate abnormal access request detection module can efficiently check out abnormal access request, and the prototype of secure storage area network system can keep high I/O performance.