Skip to Main Content
Nowadays, the growing necessity for secure and private off-site storage motivates the appearance of novel storage infrastructures. In this sense, it is increasingly common to find storage systems where users interact just with a set of' trustworthy participants, such as in Friend-to-Friend (F2F) networks. In general, these systems have been treated as a particular case of P2P networks where links among nodes are built upon social relationships. However, F2F systems are highly affected by availability correlations and very small friendsets. The combination of these phenomena poses new challenges which remain unsolved for providing an adequate storage service to users. In this work, we examine the role that correlated availabilities play in the storage service provided by a small group of friends. We extensively analyze the performance of traditional approaches used to calculate data availability and their application in the estimation of data redundancy. Moreover, we propose a novel technique to calculate the data redundancy needed based on the historical availability of a group of friends. Our study reveals important insights: i) Availability correlations in a small group of' friends can provide a good trade-off' between data availability and data redundancy. ii) Traditional redundancy calculation algorithms may importantly misestimate the data redundancy needed due to correlations. iii) Finally, our history-based algorithm exhibits high accuracy in the estimation of redundancy, providing important benefits to the system.