US 12,455,857 B2
Self-healing server file system with space cleanup
Lauren Mapp, Atlanta, GA (US)
Assigned to Truist Bank, Charlotte, NC (US)
Filed by Truist Bank, Charlotte, NC (US)
Filed on Jan. 16, 2024, as Appl. No. 18/413,344.
Application 18/413,344 is a continuation of application No. 18/061,165, filed on Dec. 2, 2022, granted, now 11,914,550.
Prior Publication US 2024/0184748 A1, Jun. 6, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 16/16 (2019.01); G06F 11/34 (2006.01); G06F 16/17 (2019.01)
CPC G06F 16/162 (2019.01) [G06F 11/3442 (2013.01); G06F 16/1727 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method, comprising:
receiving an indication of a utilization condition for a computing resource, wherein the utilization condition is a utilization of the computing resource exceeding a preset threshold;
determining, using a machine-learning model comprising a clustering algorithm, an action responsive to the utilization condition based on the utilization condition, comprising:
identifying a candidate computing resource utilization from among a plurality of candidate computing resource utilizations;
determining a similarity measure for each candidate computing resource utilization of the plurality of candidate computing resource utilizations using a property of the candidate computing resource utilization;
determining a candidate cluster including at least the candidate computing resource utilization based on the similarity measure, wherein the similarity measure is based on a clustering threshold that is based on one or more hyperparameters; and
designating the candidate cluster for the action;
outputting a first command to execute the action responsive to the utilization condition; and
determining that the utilization condition no longer exceeds the preset threshold.