International Business Machines Corporation

Mark J Walborn

Bruno Dos Santos Silva

Shawn Snodgrass

Dynamic API allocation based on data-tagging is provided. Data that is stored in a local system is parsed and normalized. One or more highly used fields is identified and tagged. A counter corresponding to each highly used field is incremented upon each reference. Upon exceeding a threshold, data is migrated to object storage. An index is created for each highly used field. A bi-directional pipeline is created between the local system and the cloud-based system. The data structure is created in object storage in the cloud-based system. Data is dynamically migrated through the pipeline from the local system to cloud-based object storage. Cloud-based system sends an API endpoint to local system. Future data accesses to local data are redirected to object storage using the API endpoint. Local system continues monitoring data utilization. Upon utilization dropping below a threshold, data accesses are redirected to local system, using the local pointer.

Growing community of inventors

Sevierville, TN, United States of America

Mark J Walborn