Distributed Wear levelling of Flash Memories

TL;DR

This paper addresses the challenge of evenly distributing write and garbage collection operations across flash memory nodes in distributed storage systems to prevent uneven wear and prolong device lifespan.

Contribution

It introduces a novel distributed wear leveling algorithm that reduces communication overhead and improves wear distribution in flash-based distributed storage systems.

Findings

The proposed algorithm effectively balances wear across flash nodes.

Simulation results show reduced message communication costs.

The approach enhances reliability and lifespan of flash storage in distributed environments.

Abstract

For large scale distributed storage systems, flash memories are an excellent choice because flash memories consume less power, take lesser floor space for a target throughput and provide faster access to data. In a traditional distributed filesystem, even distribution is required to ensure load-balancing, balanced space utilisation and failure tolerance. In the presence of flash memories, in addition, we should also ensure that the number of writes to these different flash storage nodes are evenly distributed, to ensure even wear of flash storage nodes, so that unpredictable failures of storage nodes are avoided. This requires that we distribute updates and do garbage collection, across the flash storage nodes. We have motivated the distributed wearlevelling problem considering the replica placement algorithm for HDFS. Viewing the wearlevelling across flash storage nodes as a distributed co-ordination problem, we present an alternate design, to reduce the message communication cost across participating nodes. We demonstrate the effectiveness of our design through simulation