Maximally Recoverable Codes for Distributed Storage Systems

Maximally Recoverable Codes for Distributed Storage Systems

Explore the concept of Maximally Recoverable Codes (MRC) in this 37-minute lecture by Lalitha Vadlamani from the International Institute of Information Technology, Hyderabad. Delve into the importance of efficient node repair in distributed storage systems due to increased storage capacity. Learn about locally repairable codes and their trade-off between repair locality and storage overhead. Discover three classes of MRCs based on local parity topology: MRC with locality, MRC with hierarchical locality, and Maximally Recoverable Tensor codes. Examine bounds, constructions, and open problems related to these MRC classes, gaining insights into their application in modern storage systems.