As the amount of data being generated from both commercial and private sectors has rapidly increased, so has the demand to quickly and efficiently access it. Given the scale of data and performance requirements, this demand has greatly increased the gap between existing I/O capabilities and I/O requirements. To address the challenges of large-scale, high-performance storage we have development Ceph, a distributed file system capable of scaling to petabytes of data, while sustaining very high-performance. In this talk, we will discuss the overall design of Ceph, how it has been extended for security and reliability, and how it can be leveraged to achieve very large-scale data processing. We will explore future directions for distributed storage how they can be leveraged to meet ever increasing data demands.