Building Scalable Distributed Storage Systems: Theory vs. Practice
Abstract of the talk:
With the proliferation of Internet-based applications and mobile devices, distributed storage systems such as Amazon S3 and Google Drive have become popular in the past few years by managing performance, availability and resiliency expected by consumer as well as enterprise applications. Building scalable distributed storage systems involves making careful design tradeoffs to meet user requirements at reasonable cost. This talk covers fundamentals of storage system design and explores how concepts such as CAP Theorem and error correction codes along with practical considerations such as ease of implementation and cost have shaped the evolution of distributed storage systems.