Fault detection and prediction in storage devices

摘要

A method of fault detection includes, while in normal operation: recording data corresponding to measurements of electrical current drawn during performance of a respective operation on a specified portion of a storage device; analyzing the recorded data, including determining whether one or more predefined characteristics of the recorded data meets predetermined failure criteria; and in accordance with a determination that the recorded data meets the predetermined failure criteria, performing one or more remedial actions, the one or more remedial actions including marking the specified portion as a known-bad portion. The method, optionally, includes, determining whether one or more predefined characteristics of the recorded data meets predetermined warning criteria and does not meet the predetermined failure criteria; and in accordance with a determination that the recorded data meets the predetermined warning criteria and does not meet predetermined failure criteria, performing one or more predictive actions.

主权项

1. A method of fault detection in a storage device, the method comprising:
while in normal operation:
measuring electrical current drawn during performance of a respective operation on a specified portion of the storage device and recording data corresponding to the measured electrical current drawn during performance of the respective operation on the specified portion of the storage device;analyzing the recorded data, including determining whether one or more predefined characteristics of the recorded data meets predetermined failure criteria; andin accordance with a determination that the recorded data meets the predetermined failure criteria, performing one or more remedial actions, the one or more remedial actions including marking the specified portion of the storage device as a known-bad portion.