Clustering routines for extrapolating computing resource metrics

A non-transitory computer-readable medium for extrapolating metrics for computing resources of a computing environment, comprising program code executable in at least one computing device that, when executed by the at least one computing device, directs the at least one computing device to:

• generate classes of computing resources based at least in part on a time parameter;
• for individual ones of the classes, generate clusters of the computing resources using a clustering routine, the clusters being generated such that individual ones of the computing resources in a respective one of the clusters have similar configuration parameters with respect to one another;
• determine a number of metrics required to extrapolate the metrics to other ones of the computing resources in a corresponding one of the clusters such that extrapolated metrics generated for the other ones of the computing resources satisfy a predetermined accuracy threshold, the number of metrics determined as a function of a number of the clusters generated;
• receive the number of the metrics from a client device, individual ones of the metrics corresponding to one of the computing resources in one of the clusters;
• determine a difference between respective ones of the number of the metrics as received and a corresponding metric value stored in a reference library;
• determine an average of the difference determined for individual ones of the computing resources;
• and
• update a corresponding one of the extrapolated metrics generated for other ones of the computing resources in the one of the clusters using the individual ones of the metrics and the corresponding metric value stored in the reference library based at least in part on the average of the difference.

A computer-implemented method for extrapolating metrics for computing resources of a computing environment, comprising:

• generating classes of computing resources based at least in part on a time parameter;
• for individual ones of the classes, generating clusters of the computing resources using a clustering routine, the clusters being generated such that individual ones of the computing resources in a respective one of the clusters have similar configuration parameters with respect to one another;
• determining a number of metrics required to extrapolate the metrics to other ones of the computing resources in a corresponding one of the clusters such that extrapolated metrics generated for the other ones of the computing resources satisfy a predetermined accuracy threshold, the number of metrics determined as a function of a number of the clusters generated;
• receiving the number of the metrics from a client device, individual ones of the metrics corresponding to one of the computing resources in one of the clusters;
• determining a difference between respective ones of the number of the metrics as received and a corresponding metric value stored in a reference library;
• determining an average of the difference determined for individual ones of the computing resources;
• and
• updating a corresponding one of the extrapolated metrics generated for other ones of the computing resources in the one of the clusters using the individual ones of the metrics and the corresponding metric value stored in the reference library based at least in part on the average of the difference.