New Replication Strategy Based on Maximal Frequent Correlated Pattern Mining for Data Grids

Abstract

Data replication in data grids is an efficient technique that aims to improve response time, reduce the bandwidth consumption and maintain reliability. In this context, a lot of work is done and many strategies have been proposed. Unfortunately, most of existing replication techniques are based on single file granularity and neglect correlation among different data files. Indeed, file correlations become an increasingly important consideration for performance enhancement in data grids. In fact, the analysis of real data intensive grid applications reveals that job requests for groups of correlated files and suggests that these correlations can be exploited for improving the effectiveness of replication strategies. In this paper, we propose a new dynamic periodic decentralized data replication strategy, called RSBMFCP ( 1 ) , which consider a set of correlated files as granularity. Our strategy gathers files according to a relationship of simultaneous accesses between files by jobs and stores correlated files at the same site. In order to find out these correlations, a maximal frequent correlated pattern mining algorithm of the data mining field is introduced. We choose the all-confidence as correlation measure. The proposed strategy consists of four steps: storing file access history, converting the file access history into a logical history file, applying maximal frequent correlated pattern mining algorithm and performing replication and replacement. Experiments using the well-known data grid simulator OptorSim show that our proposed strategy has better performance in comparison with other strategies in terms of job execution time and effective network usage.

A Critical Survey of Data Grid Replication Strategies Based on Data Mining Techniques

“…Indeed, the resulting set of association rules has an excessively large size, with a majority of the mined rules either redundant or do not reflect the true correlation relationship among data [19]. The RSBMFCP [27] tries to overcome this limitation by proposing the use of the all-confidence correlation measure in the data mining process. A second observation concerns the data used to infer file correlations.…”

“…Slimani et al proposed in [27] a new strategy, called RSBMFCP, which is composed by four steps: (i) extracting file access history. (ii) converting the file access history into an extraction context (i.e., a logical file access history).…”

A data mining correlated patterns-based periodic decentralized replication strategy for data grids

“…Firstly, we survey in detail the main replication strategies which take into account file correlations, and more specifically, those based on data mining techniques. Compared to [57], a critical discussion of these strategies is then carried out in this extended version. This will allow us to highlight the limitations of the strategies of the literature which motivate our proposal.…”

“…Replication is based on the granularity of a set of correlated files. In comparison to [57], further details are…”

“…Thirdly, we conduct an in-depth comparison with other replication strategies. In [57], we have compared our strategy with other replication strategies, namely, No Replication, DR2 [58], Periodic Optimiser [7] and DPRS-KP [12]. The carried out experiments showed that our approach gives better performance w.r.t.…”