Systems and methods for edge system resource capacity performance prediction
Managing the resource demand load for edge systems is significantly more complex than for other systems, such as cloud environments. Unlike cloud systems and other frameworks that are able to use closed-form solutions based on Poisson processes or other tractable Gaussian-based probability distributions, edge systems present complex waveforms, pareto/alpha-stable distributions, and long-range dependence. Based on elaborately designed embodiments that recognize the complexities of edge data, one can estimate scaling and multi-fractal dimensionality to determine predictive models.
1 . A processor-implemented method comprising:
determining resource capacity of a resource of an edge system using a multi-fractal dimensional methodology;
given the determined resource capacity, using a queuing-type model to determine a forecasted resource demand load that can be served by the edge system at a given blocking percentage;
using a resource uncertainty estimation (RUE) methodology on resource-related data collected from the edge system to obtain a short-term resource demand load;
comparing the short-term resource demand load to the forecasted resource demand load to identify a disparity between the short-term resource demand load and the forecasted resource demand load that indicates a potential issue; and
responsive to detecting a disparity between the short-term resource demand load and the forecasted resource demand load, causing one or more actions to be taken to eliminate or mitigate the potential issue.
2 . The processor-implemented method of claim 1 wherein the step of determining resource capacity of a resource of an edge system using a multi-fractal dimensional methodology comprises:
collecting time series of data from the edge system;
using one or more wavelets to analyze the time series of data to determine a set of one or more Hurst parameters;
using a regression analysis of the time series of data to obtain parameters for an objective function; and
given the one or more Hurst parameters and the parameters for the parameters from the objective function, estimating resource capacity versus demand resource load for edge sites of the edge system.
3 . The processor-implemented method of claim 2 wherein the step of collecting time series of data from the edge system comprises:
collecting data at a first time interval; and
summarizing a set of data collected at the first time interval to form a set of data for the time series of data.
4 . The processor-implemented method of claim 1 wherein the step of using a resource uncertainty estimation (RUE) methodology on resource-related data collected from the edge system to obtain a short-term resource demand load comprises:
using a dataset comprising the resource-related data and an M-PCM-OFFD (Multivariate Probabilistic Collocation Method-Orthogonal Fractional Factorial Design) methodology to obtain the short-term resource demand load.
5 . The processor-implemented method of claim 1 wherein the one or more actions comprise one or more of:
increasing blocking probability for the edge system;
adding capacity to one or more edge sites of the edge system;
differentiating between service levels for tasks; and
sending a task to another edge system to be handled.
6 . The processor-implemented method of claim 1 wherein the queuing-type model is a modified Erlang distribution model.
7 . The processor-implemented method of claim 1 wherein the forecasted resource demand load and the short-term resource demand load are performed at a resource level for an edge site.
8 . One or more information handling systems collectively comprising:
one or more processors; and
one or more non-transitory computer-readable media comprising one or more sets of instructions which, when executed by at least one of the one or more processors, causes steps to be performed comprising:
determining resource capacity of a resource of an edge system using a multi-fractal dimensional methodology;
given the determined resource capacity, using a queuing-type model to determine a forecasted resource demand load that can be served by the edge system at a given blocking percentage;
using a resource uncertainty estimation (RUE) methodology on resource-related data collected from the edge system to obtain a short-term resource demand load;
comparing the short-term resource demand load to the forecasted resource demand load to identify a disparity between the short-term resource demand load and the forecasted resource demand load that indicates a potential issue; and
responsive to detecting a disparity between the short-term resource demand load and the forecasted resource demand load, causing one or more actions to be taken to eliminate or mitigate the potential issue.
9 . The one or more information handling systems of claim 8 wherein the step of determining resource capacity of a resource of an edge system using a multi-fractal dimensional methodology comprises:
collecting time series of data from the edge system;
using one or more wavelets to analyze the time series of data to determine a set of one or more Hurst parameters;
using a regression analysis of the time series of data to obtain parameters for an objective function; and
given the one or more Hurst parameters and the parameters for the parameters from the objective function, estimating resource capacity versus demand resource load for edge sites of the edge system.
10 . The one or more information handling systems of claim 9 wherein the step of collecting time series of data from the edge system comprises:
collecting data at a first time interval; and
summarizing a set of data collected at the first time interval to form a set of data for the time series of data.
11 . The one or more information handling systems of claim 8 wherein the step of using a resource uncertainty estimation (RUE) methodology on resource-related data collected from the edge system to obtain a short-term resource demand load comprises:
using a dataset comprising the resource-related data and an M-PCM-OFFD (Multivariate Probabilistic Collocation Method-Orthogonal Fractional Factorial Design) methodology to obtain the short-term resource demand load.
12 . The one or more information handling systems of claim 8 wherein the one or more actions comprise one or more of:
increasing blocking probability for the edge system;
adding capacity to one or more edge sites of the edge system;
differentiating between service levels for tasks; and
sending a task to another edge system to be handled.
13 . The one or more information handling systems of claim 8 wherein the queuing-type model is a modified Erlang distribution model.
14 . The one or more information handling systems of claim 8 wherein the forecasted resource demand load and the short-term resource demand load are performed at a resource level for an edge site.
15 . A non-transitory computer-readable medium or media comprising one or more sequences of instructions which, when executed by at least one processor, causes steps to be performed comprising:
determining resource capacity of a resource of an edge system using a multi-fractal dimensional methodology;
given the determined resource capacity, using a queuing-type model to determine a forecasted resource demand load that can be served by the edge system at a given blocking percentage;
using a resource uncertainty estimation (RUE) methodology on resource-related data collected from the edge system to obtain a short-term resource demand load;
comparing the short-term resource demand load to the forecasted resource demand load to identify a disparity between the short-term resource demand load and the forecasted resource demand load that indicates a potential issue; and
responsive to detecting a disparity between the short-term resource demand load and the forecasted resource demand load, causing one or more actions to be taken to eliminate or mitigate the potential issue.
16 . The non-transitory computer-readable medium or media of claim 15 wherein the step of determining resource capacity of a resource of an edge system using a multi-fractal dimensional methodology comprises:
collecting time series of data from the edge system;
using one or more wavelets to analyze the time series of data to determine a set of one or more Hurst parameters;
using a regression analysis of the time series of data to obtain parameters for an objective function; and
given the one or more Hurst parameters and the parameters for the parameters from the objective function, estimating resource capacity versus demand resource load for edge sites of the edge system.
17 . The non-transitory computer-readable medium or media of claim 16 wherein the step of collecting time series of data from the edge system comprises:
collecting data at a first time interval; and
summarizing a set of data collected at the first time interval to form a set of data for the time series of data.
18 . The non-transitory computer-readable medium or media of claim 15 wherein the step of using a resource uncertainty estimation (RUE) methodology on resource-related data collected from the edge system to obtain a short-term resource demand load comprises:
using a dataset comprising the resource-related data and an M-PCM-OFFD (Multivariate Probabilistic Collocation Method-Orthogonal Fractional Factorial Design) methodology to obtain the short-term resource demand load.
19 . The non-transitory computer-readable medium or media of claim 15 wherein the one or more actions comprise one or more of:
increasing blocking probability for the edge system;
adding capacity to one or more edge sites of the edge system;
differentiating between service levels for tasks; and
sending a task to another edge system to be handled.
20 . The non-transitory computer-readable medium or media of claim 15 wherein the queuing-type model is a modified Erlang distribution model.