Paper Title: A Global-Scale Cloud Storage Platform Based on Nonvolatile Memory

Paper Abstract: A global-scale cloud storage platform must be able to store exabytes of customer data in multiple forms and serve millions of customer requests per second. It should be designed for high performance customized flash-based SSDs and emerging non-volatile memory such as 3D XPoint. A storage engine implemented at Alibaba includes a KV engine, user space filesystem, and customized driver, which are highly optimized for the storage hardware (customized flash-based SSD and 3D XPoint). The platform demonstrates a balance between performance and system cost that is capable of serving thousands of different types of applications efficiently.

Paper Author: Fei Liu, Storage Architect, Alibaba

Author Bio: Fei Liu is a Storage Architect at Alibaba, where he designs and deploys the next-generation storage system for the company’s public cloud and e-commerce businesses. Before joining Alibaba, he was an architect at Samsung where he designed next-generation SSDs and software-defined storage systems for data centers. He has also worked at LSI, SanDisk, and Western Digital. A 14-year storage industry veteran, he holds three patents. He earned a Master's degree in computer science at the University of Electronic Science and Technology of China.