Abstract
In large-scale data storage systems, disk failures are common. Erasure codes are employed to store data in a redundant fashion to protect against data loss. In this talk, I will present
(1) reasons for why adapting the code parameters over time can lead to substantial resource savings,
(2) present a theoretical framework to formalize the notion of "code conversion"---the process of converting data encoded using one set of code parameters into data encoded using a different set of code parameters, while maintaining desired decodability properties, and
(3) introduce "convertible codes", a new class of codes that enable resource-efficient conversion, bounds on their resource requirements, and explicit constructions.