Abstract: In order to solve the problem of high power consumption and low efficiency of information storage caused by the enlargement of large data environment, a data cloud storage optimization algorithm based on LSM (Log Structured Merge) tree is proposed. According to the difference of the quantity and size of data storage, network bandwidth and link length, the judgement model of data fragmentation storage is established, and the time delay of data sending and receiving is calculated. For the data that need to be stored in segments, the server power consumption, static power consumption and dynamic power consumption under each node are determined by the time delay threshold, and the data with high average power consumption are transmitted by category to achieve storage and energy saving optimization. Simulation results show that the proposed method can reduce the amount of redundant data in cloud storage environment, and the processing stability is strong, the average consumption is lower than the set threshold.

Key words: log structured merge ( LSM) tree; , network bandwidth; , link length; , static power consumption; , redundant data