时间：6月15日下午15:00
地点：东五楼210

Abstract:
Various key-value stores are widely employed for data management to support Internet services as they offer higher efficiency,scalability,and availability than relational database system. The log-structured merge tree based KV stores have attracted growing attention because they can eliminate random writes and maintain acceptable read performance. Recently, as the price per unit capacity of NAND flash decreases, solid state disks have been extensively adopted in enterprise-scale data centers to provide high I/O bandwidth and low access laency. However, iti is inefficient to naively combine LSM-tree-based KV stores with SSDs, as the high parallelism enabled within the SSD cannot be fully exloited. Current LSM-tree-based KV stores are designed without asuming SSD's multi-channel architecture. To address this inadequacy, we propose LOCS, a system equipped with a customized SSD design, which exposes its internal flash channels to applications, to work with the LSM-tree-based KV store, specifically LevelDB in this work. We extend LevelDB to explictly leverage the multiple channels of an SSD to exploit its abundant parallelism. In addition, we optimize scheduling and dispatching polices for concurrent I/O requests to further improve the efficiency of data access.

Bio:
Song Jiang received the Ph.D in Computer Science, College of William and Mary in June 2004. He currently is an associate professor at The University of Texas at Arlington.His many research results have been adopted by well-known open source communies. Token-Ordered LRU algorithm has been officily adopted in the Linux Kernel since version 2.6. Linux that is the most popular open-source operating system. The Clock-pro replacement algorithm has been officially adopted in the NetBSD kernel. The LIRS replacement algorithm has been officially adopted in MySQL. MySQL is the world's most popular open source database software used by Google and Wikipedia. Dr.Song Jiang is engaged in System infrastructure for big data, in particular, data management systems, file and storage systems, and I/O performance in high-performance computing, and has published more than 70 papers in international conferences and journals. His research was funded by the National Natural Science Foundation of the United States and the Argonne National Laboratory.