US 11,809,267 B2
Root cause analysis of computerized system anomalies based on causal graphs
Mircea R. Gusat, Langnau am Albis (CH); Lili Lyubchova Georgieva, Sofia (BG); Serge Monney, Pully (CH); and Charalampos Pozidis, Thalwil (CH)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed on Apr. 8, 2022, as Appl. No. 17/658,483.
Prior Publication US 2023/0325269 A1, Oct. 12, 2023
Int. Cl. G06F 11/07 (2006.01)
CPC G06F 11/079 (2013.01) [G06F 11/0709 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method of root cause analysis of computerized system anomalies, wherein the method comprises:
monitoring key performance indicators (KPIs) for a computerized system of interest, wherein KPI values of the monitored KPIs form respective timeseries;
detecting an anomaly in the computerized system based on the monitored KPIs;
determining a troubleshooting time window extending over a given time period, in accordance with the detected anomaly;
identifying a strict subset of the monitored KPIs based on portions of the respective timeseries spanning the given time period, wherein the strict subset comprises abnormal KPIs (aKPIs) and potential explanatory KPIs (xKPIs);
obtaining a causal graph of vertices mapping KPIs of the strict subset by running a causality algorithm to evaluate weights of directed edges connecting the vertices and accordingly obtain one or more directed paths, each connecting one of the xKPIs to one of the aKPIs; and
returning the obtained causal graph to help troubleshoot the detected anomaly.