US 11,991,059 B1
Techniques for generating service-specific server health scores and application traffic forecasts
Prerana Dharmesh Gambhir, San Jose, CA (US); Sharena M. Pari-Monasch, Union City, CA (US); Qiong Zhou, Cupertino, CA (US); Thanh Trung Nguyen, Bellevue, WA (US); Sarah Ferraro Stein, Sunnyvale, CA (US); Christine Bumpous, Renton, WA (US); and Daniel M. Cheung, San Francisco, CA (US)
Assigned to Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed by Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed on Dec. 16, 2022, as Appl. No. 18/083,255.
Int. Cl. G06F 15/173 (2006.01); H04L 41/16 (2022.01); H04L 43/045 (2022.01); H04L 43/062 (2022.01); H04L 43/067 (2022.01)
CPC H04L 43/067 (2013.01) [H04L 41/16 (2013.01); H04L 43/045 (2013.01); H04L 43/062 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
for each server computer in a predefined group of server computers executing an instance of a first service, i) obtaining for the server computer a value for each of a plurality of utilization metrics, and ii) providing, as input features, the value of each of the plurality of utilization metrics to a first pre-trained machine learning model to derive, as output, a service-specific server health score for the server computer;
providing, as input features, a measure of application traffic processed by the predefined group of server computers for each of a plurality of prior time periods to a second pre-trained machine learning model to derive, as output, a measure of predicted application traffic to be received by the predefined group of server computers in a future time period;
processing one or more rules, by a rules-based engine, using as input to the rules-based engine the service-specific server health score for each server computer in the group of server computers executing the instance of the first service and the measure of predicted application traffic;
presenting, via a user interface, a recommendation relating to the future time period, the recommendation derived at least in part by a result of the processing of the one or more rules; and
receiving user input via the user interface to invoke an operation consistent with the recommendation relating to the future time period, the operation comprising one of:
configuration of the first service on one or more additional server computers;
termination and removal of the first service from one or more server computers; or
reconfiguration of application traffic routing to the predefined group of server computers.