Aims and Motivation The goal of this project is to produce a secure and dependable way of distributing and storing data securely over a distributed system while also being able to tolerate faults within the system. Faults should not affect the ability of the system to recover information previously distributed. Dependability of the system is based on the ability to avoid service failures that are more frequent and more severe than is acceptable to the operation of the system. Security is a composite of the attributes of confidentiality, integrity and availability [1]. Methods The distributed system contains both client and storage nodes while employing: a) Cryptographic hash functions or digests, like MD5 and SHA-1,are used in: the “unique” naming of fragments fragment data integrity checking b) Symmetric block-cipher algorithms, like Advanced Encryption Standard (AES) and Digital Encryption Standard (DES), to encrypt the raw data to be fragmented; c) Fragmentation Redundancy and Scattering (FRS) technique is used for distributing fragments redundantly within the system, improving the reliability and availability of the system. Computing Laboratory Implementation The Aglet System was used to implement and simulate the prototype FRS system. Aglets are the Java software instantiation of an agent. Each Aglet is a persistent object, that lives and interacts within the Aglet system and can travel to other Aglet systems, storing state and processing information until actively disposed of. A Message object infrastructure facilitates a multicasting service communication between aglets. Client Aglets requested services from Servers for the storage and retrieval of fragments. Server storage nodes stored the fragments. Results Preliminary results produced suggest that a large number of fragments created from the fragmentation of a large data source could overload or overwhelm the system. More results need to be produced to make a valid comment on how to change the present system to meet the requirements. Conclusion The system produced proposes the framework for a possible FRS-based distributed storage system. Dependability is ensured through the redundancy of fragments stored within the system. Security is achieved through the encryption, random scattering and random retrieval of fragments. The cost of the system is, however, the large communication overhead relating to the scattering and retrieval of fragments. Future Work Future work will aim mainly at tackling problems such as adding fault to the system and schemes for optimising the system. References 1.Avizienis, A., Laprie, J., Randell, B., Landwehr, C. (2004). "Basic Concepts and Taxonomy of Dependable and Secure Computing," IEEE Transactions on Dependable and Secure Computing, vol. 01, no. 1, pp , January-March, Y. Deswarte, L. Blain, J.-C. Fabre. “Intrusion Tolerance in Distributed Computing Systems”. Proceedings of the IEEE Symposium on Security and Privacy. Oakland, California, USA. May pp R. Ball, V. Spurrett, R. de Lemos. “Dependable and Secure Storage for Multi-Agent Systems”. 5th International Workshop of Software Engineering for Large-scale Multi-Agent Systems (SELMAS). Shanghai, China. Submission February Tolerating Faults in a Distributed Storage System Rudi Ball, Computing Laboratory – University of Kent, UK Further Details Rudi Ball