Towards automated personalized data storage

User data is growing at an ever greater pace that threatens to overwhelm our ability to effectively manage it. As the types of data increase, and the storage environments become ever more heterogeneous, even reasoning about basic data management decisions becomes increasingly difficult. This expansion in complexity requires new methodologies for managing data that alleviate as much of the burden as possible from the individual user. Instead of requiring users to understand their full collection of data and the underlying storage architectures, future storage systems need to be able to decide on their own how to manage individual files both in terms of the appropriate storage medium as well as the necessary file operation semantics. In this paper we present a vision for future storage systems that address the dramatic increase in complexity and volume by providing autonomic storage management decisions based on dynamically collected metrics that measure the relationship between individual users and each of their personal files.