The Study Of Wear-aware Memory Management Scheme For Multiple Non-volatile Memory Slots

Recently,with the rapid development of technologies such as cloud computing,big data,artificial intelligence,smart city and transportation,and automatic driving,it has caused large-scale expansion of digital resources.The performance requirements of these technologies requires the storage capacity and computing power of the underlying computer system to support.Traditional storage systems typically use disks as the underlying persistent storage device.However,due to the huge performance gap between disk and memory,the performance of disk-based storage systems is limited.Emerging Non-Volatile Memory(NVM)has many advantages,such as near-DRAM speed,byte-addressability and persistence.Modern computer systems contain many memory slots,which are exposed as a unified storage interface by a shared address space.Since NVM has limited write endurance,many wear-leveling techniques are implemented in hardware.However,existing hardware techniques can only effective in a single NVM slot,which cannot ensure wear-leveling among multiple NVM slots.This paper explores how to optimize a storage system with multiple NVM slots in terms of performance and lifetime.This paper shows the huge gap in the distribution of write data in multiple NVM slots by studying existing file systems based on NVM.Thus reveals that simple integration of multiple NVMs in traditional memory policies results in poor reliability and existing hardware wear-leveling technologies are ineffective for a system with multiple NVM slots.In this paper,we propose a common wear-aware memory management scheme for in-memory file system.The proposed memory scheme enables wear-aware control of NVM slot use which minimizes the cost of performance and lifetime.Moreover,we provide three wear-leveling allocation strategies taking into account different workload pattern.The proposed three allocation strategies coupled with wear counter to balance the write times and improve both lifetime and performance for different concurrent workload pattern in a system with multiple NVM slots.We implemented the proposed memory management scheme and evaluated their effectiveness.The experiments show that the proposed wear-aware memory management scheme can outperform wearleveling effect by more than 478 x,and the lifetime of NVM can be prolonged by 2.5x,the performance can be improved by up to 15%.And from the experimental results,wesummarize a decision table to guide us to use which allocation strategy is the most appropriate.