TSUE: A Two-Stage Data Update Method for an Erasure Coded Cluster File System

Erasure coding is widely adopted by large-scale storage providers to reduce data maintenance costs. However, compared to replication storage systems, erasure coded storage introduces significantly more overhead during data updates. To address this issue, researchers have proposed various parity logging methods to improve update efficiency. Nevertheless, due to the significant amount of random I/O involved in erasure code update processes, the long latency and low throughput often fail to meet the requirements of online storage systems. To this end, we propose a two-stage data update method called TSUE. TSUE divides the update process into a synchronous stage that records updates in a data log, and an asynchronous stage that recycles the data log later. This approach effectively transforms the random I/O generated by updates into sequential I/O through the recording in the data log, and it merges updates to reduce the computing overhead of erasure coding through recycling the data log. In SSDs cluster, TSUE significantly improves update performance, achieving improvements of 7.6× under Ali-Cloud trace, 5× under Ten-Cloud trace, while it also significantly enhances SSD lifespan by reducing erase writing by up to 90%.