Performance Analysis of Centralized versus Distributed Recovery Schemes in P2P Storage Systems

Abstract

Abstract. This paper studies the performance of Peer-to-Peer Storage Systems (P2PSS) in terms of data lifetime and availability. Two schemes for recovering lost data are modeled through absorbing Markov chains and their performance are evaluated and compared. The first scheme relies on a centralized controller that can recover multiple losses at once, whereas the second scheme is distributed and recovers one loss at a time. The impact of each system parameter on the performance is evaluated, and guidelines are derived on how to engineer the system and tune its key parameters in order to provide desired lifetime and/or availability of data. We find that, in stable environments such as local area or research laboratory networks where machines are usually highly available, the distributed-repair scheme offers a reliable, scalable and cheap storage/backup solution. This is in contrast with the case of highly dynamic environments, where the distributed-repair scheme is inefficient as long as the storage overhead is kept reasonable. P2PSS with centralized-repair scheme are efficient in any environment but have the disadvantage of relying on a centralized authority. Our analysis also suggests that the use of large size fragments reduces the efficiency of the recovery mechanism.

A realistic simulation model for peer-to-peer storage systems

“…As a result, our simulations suggest that the models presented in [2] give accurate results on data durability and availability only in replicated P2PSS (as in [20]) or in Experiment 4 like-context. This was the motivation to realize the work in [5] where the models evaluate data durability and availability in more wide contexts.…”

“…The resulting models, appeared in [5], characterize data lifetime and availability in P2PSS storage systems that use either replication or erasure codes, under more realistic assumptions and well-known contexts as supported by results of the simulator presented in this paper and in [6].…”

Lifetime and availability of data stored on a P2P system: Evaluation of redundancy and recovery schemes

“…In light of the conclusions of [28], namely, that machine availability is modeled with a hyper-exponential distribution, and building on the findings of [12] we will propose in this paper a more general and accurate models that are valid under different distributed environments. Thus, the models presented in Sections 5 and 6 can be seen as a generalization of those presented in [10] and has little in common with [1]. Although [1,10] are not valid under different systems and can not give a very accurate results, they are very simple and give a good background to easiely understand this paper.…”

“…Thus, the models presented in Sections 5 and 6 can be seen as a generalization of those presented in [10] and has little in common with [1]. Although [1,10] are not valid under different systems and can not give a very accurate results, they are very simple and give a good background to easiely understand this paper.…”

“…Given Assumptions 1-4, the models developed in this report are more general and/or more realistic than those in [30,1,10]. Table 1 recapitulates the parameters introduced in this section.…”