Transmission Layer ( Byte Addressable )

Virtualization-based cloud computing has dominated today’s data centers by supporting consolidated servers, converged infrastructures, horizontal scalability, and cost of ownership reduction. As the demands of cloud-based services continue to grow, numerous enabling technologies have been developed such as system virtualization (e.g., VM, container and NFV), software-defined management (for computing, networking and storage), and pyramid service stacks (e.g., SaaS, PaaS and IaaS). Next-generation hardware (e.g., nonvolatile memories and high-speed networks) further fires up the evolvement of these cloud-based solutions. Correspondingly, new research challenges arise from the cloud computing paradigm, and I am passionate about developing practical, systematic solutions to address these challenges. My research interests span operating systems/hypervisors, distributed systems, and performance analysis and optimization, especially in the context of virtualization-based cloud infrastructures. My Ph.D. research centers around file and storage systems, and involves building fair, reliable and efficient storage systems to empower virtualized clouds. Specifically, in consideration of new applications and technology shifts, my research re-visited established cloud systems, identified critical-but-overlooked system deficits, and redesigned and implemented cloud-compatible systems. I have designed and implemented an integrated cloud storage framework with high I/O performance [1, 3], fair resource allocation/scheduling [2], and enhanced data security and reliability [4]. I have also worked on other cloud-related problems, including defining an optimal multiple-VM migration scheme [5], developing an intelligent SDN-based controlling system for hybrid networks [6], and performing large-scale workload characterization and performance analysis on massive, multi-core systems [7, 8]. My research efforts reflect the following characteristics. First, I believe re-examining established computing systems in response to new software, hardware and applications is key to identifying new, realistic system research challenges and opportunities. For instance, my work [2] exposes unfairness among competing I/O flows with a thorough reexamination of storage I/O behaviors under virtualization. Second, I believe that building prototypes on real software/hardware infrastructures is a crucial way to help system researches understand real challenges behind their proposed ideas, as well as to discover hidden design pitfalls. During my prototyping effort towards a cloud storage middle-box system [4], I observed an I/O data granularity mismatch problem between storage and network layers in a multi-layered storage infrastructure. This observation led to another research idea and solution, in which I re-architected the cloud block storage systems [1]. Third, my past research projects in various systems areas such as virtualization, storage, and software-defined networking give me a full spectrum of experience, insights and curiosity to explore new research opportunities in systems. Finally, many research problems I address come directly from my substantial collaborations with industry; my connections with industry labs (those at Microsoft, AT&T, Intel, and NEC) give me unique opportunities to access real-world cloud infrastructures, workloads, and traces; obtain feedback from practitioners; and evaluate my solutions and prototypes in realistic settings. In the following, I will first describe in detail my Ph.D. research in cloud storage systems, and then briefly discuss my other research projects in cloud computing. Finally, I will present my future research plan. Figure 1 illustrates my Ph.D. research efforts (in red) and future research ideas (in blue), and shows how they relate to each other.