Economic Modeling of Long-Term Storage
As society’s infrastructure continues to transition to digital form, reliable preservation of digital data becomes increasingly important, and economic considerations are key to this effort. Unlike physical objects, digital objects require intervening technology to access, and that intervening technology is typically replaced multiple times throughout the life of a digital object. Since technological progress is largely unpredictable, so is the cost of technology replacement cycles. This places a burden on archive owners to either absorb unforeseen costs directly – long after the data to which the costs can be attributed to has been stored – or, more likely, to under-provision the storage system, saving costs but endangering data. For long-lived data, this situation will likely arise not just once, but repeatedly throughout the lifetime of the data. We develop a simulation model of a long-term storage system and study the relationships between various technological factors and long-term cost. Our aim is to develop a deeper understanding of the fundamental trade-off between monetary cost of storage and longevity of digital data.
Presenter: Daniel Rosenthal, Fourth Year Ph.D. Student, University of California, Santa Cruz