System and method for reduced SSD failure via analysis and machine learning
AssigneeKioxia Corporation
InventorsYaron Klein, Verly Gafni-Hoek
Life of this patent
Estimating the term…
Abstract
Various implementations described herein relate to systems and methods for predicting and managing drive hazards for Solid State Drive (SSD) devices in a data center, including receiving telemetry data corresponding to SSDs, determining future hazard of one of those SSDs based on an a-priori model or machine learning, and causing migration of data from that SSD to another SSD.
Claims20
A central management device configured to manage a plurality of solid state drive (SSD) devices in a data center, the central management device comprising:
- a prediction module implemented with a processing circuit having a processor and a memory, the prediction module configured to:
- identify drive hazard information;
- identify a drive model that is generated by using machine learning to determine correlations between telemetry information and the drive hazard information, wherein the drive model is configured to predict future drive hazards;
- and
- determine, based on the drive model and current telemetry information, a future hazard of a first SSD device of the plurality of SSD devices;
- and
- a drive assignment module that is configured to manage storage of data in the plurality of SSD devices based on determinations by the prediction module.
A method for managing a plurality of solid state drive (SSD) devices in a data center, the method comprising:
- identifying drive hazard information corresponding to the plurality of SSD devices;
- identifying a drive model that is generated by using machine learning to determine correlations between telemetry information and the drive hazard information, wherein the drive model is configured to predict future drive hazards;
- and
- determining, based on the drive model and telemetry information, future hazard of a first SSD device of the plurality of SSD devices.
A central management device configured to manage a plurality of solid state drive (SSD) devices in a data center, the central management device comprising:
- a prediction module implemented with a processing circuit having a processor and a memory, the prediction module configured to:
- determine, based on an a-priori drive model, data I/O workload information and a current drive state, a predicted drive state for each of the plurality of SSD devices;
- determine differences between the predicted drive state and an actual drive state;
- and
- update, based on the differences, the a-priori drive model;
- and
- a drive assignment module that is configured to manage storage of data in the plurality of SSD devices based on determinations by the prediction module.
Dependencies are inferred from the claim text.