Repairable Data Blocks and Method of Data Block Reconstruction for Large Scale Cloud Storage Systems

Case ID:

This invention is a new approach to reinforcing the storage system by storing the data with specifically placed redundancies that reduce the chance that all copies of the same data will be lost simultaneously. The invention also allows for possible reconstruction of lost data using remaining data.



Online cloud computing systems utilizing distributed data storage actually divide the information into “blocks” which are stored amongst several different nodes. If and when some of these nodes fail, the blocks they are storing can be at least temporarily lost, causing the aggregate information to become fragmented. Usually this is remedied by storing additional copies as a failsafe, but this can be costly.



The current invention can be applied to any cloud computing system relying on distributed data storage, and can be expected to increase its efficiency, and therefore capacity.



This invention has been demonstrated to be substantially superior to the state-of-the-art in theory, as there is little doubt that real-world usage would show significant improvements as compared to traditional methods.

  • Proposed method reduces storage dedicated to redundancy required to maintain resilience
  • Proposed method increases failure resilience
  • Proposed method is not more costly than any other method
Patent Information:
Contact For More Information:
Tariq Ahmed
Sr Licensing Manager, College of Engineering
The University of Arizona
Lead Inventor(s):
Onur Koyluoglu
Gokhan Calis