Optimization And Implementation Of Data Layout Strategy Based On Local Reconstruction Code

With the continuous development of science and technology,the era of big data has quietly come to the fore,and the huge amount of data has brought enormous challenges to the storage system.With the continuous expansion of the scale of distributed storage systems,cheap devices are put into use,and the failure of storage systems is becoming more frequent.How to ensure the high reliability of data has become a key issue to be solved.Erasure code is a very common fault-tolerant mechanism in distributed storage systems.And it can minimize the storage overhead while ensuring the same fault tolerance,and thus has become a research hotspot.At present,most researches on error correction of erasure codes are based on the node level,and there is no further optimization for the disk composition inside the nodes.The data layout strategy based on local repair code is a data layout strategy for multi-disk nodes.The data layout is designed with the disk as the minimum unit,and the fault tolerance of the disk is increased.At the same time,the repair performance of the local reconstruction code is optimized,and the data blocks required for the repair process are concentrated on as few nodes as possible,thereby reducing the transmission overhead during data repair.This enables the storage system to meet both node-level fault tolerance and better repair performance in the event of a single disk failure.After the theoretical analysis,the encoding was implemented in an IEC system.The performance of the IEC system was tested using the R-LRC code and the 3 copy strategy as the control group.The test results show that when using local repair code technology for fault tolerance,the data layout strategy based on local repair code has higher disk fault tolerance.Compared with the traditional random layout strategy,it has better data repair performance,and also ensures better write performance and read performance.