# US 2022/0035668 A1 · Resource management of a software application with multiple software components

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2022/0035668 A1
- Document ID: US20220035668A1
- Status: Published application
- Kind: A1
- Application number: 16/944,197
- Filed: 2020-07-31
- Published: 2022-02-03
- Earliest claimed: 2020-07-31
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Gabriele de Capoa; Massimo Villani
- Web page: https://patentgrep.com/patent/US20220035668A1

## CPC

- G06F 9/505: {considering the load}
- G06F 11/302: {where the computing system component is a software system}
- G06F 11/3055: {Monitoring arrangements for monitoring the status of the computing system or of the computing system component, e.g. monitoring if the computing system is on, off, available, not available (error or fault processing without redundancy G06F11/0703; error detection or correction by redundancy in data representation G06F11/08; error detection or correction of the data by redundancy in operations G06F11/14; error detection or correction by redundancy in hardware G06F11/16)}
- G06F 11/3433: {for load management (allocation of a server based on load conditions G06F9/505; load rebalancing G06F9/5083; redistributing the load in a network by a load balancer H04L67/1029)}
- G06F 11/3442: {for planning or managing the needed capacity}
- G06F 11/3447: {Performance evaluation by modeling}
- Y02D 10/00: Energy efficient computing, e.g. low power processors, power management or thermal management
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- Y02D: CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE

## Abstract

A solution is proposed for resource management of a software application including a plurality of software components interacting with each other. A corresponding method includes monitoring present conditions of the software components and estimating a future consumption of one or more computing resources by each software component from the present conditions of the software components; an allocation of the computing resources to the software components is then controlled accordingly. A computer program and a computer program product for performing the method are also proposed. Moreover, a system for implementing the method is proposed.

## Claims

### Claim 1

A method for resource management of a software application comprising a plurality of software components interacting with each other, wherein the method comprises:
monitoring, by a computing system, corresponding present conditions of the plurality of software components;
estimating, by the computing system for each of the plurality of software components, corresponding future consumptions of one or more computing resources each indicative of a future value of a consumption of the corresponding computing resource by the software component, the future consumptions of the software component being estimated from the present conditions of the software components according to at least one estimation model; and
controlling, by the computing system, an allocation of the computing resources to the software components according to the corresponding future consumptions.

### Claim 2 (depends on 1)

The method according to claim 1, wherein the method comprises:
monitoring, by the computing system for each of the plurality of software components, the present condition of the software component comprising corresponding present metrics of one or more workload metrics of the software component each indicative of a present value of a corresponding workload metric.

### Claim 3 (depends on 1)

The method according to claim 1, wherein said at least one estimation model comprises an application model of the software application and corresponding component models of the plurality of software components, the method comprising:
estimating, by the computing system, corresponding future conditions of the plurality of software components, the future condition of each of the software components being estimated from the present conditions of the software components according to the application model; and
estimating, by the computing system for each of the plurality of software components, the future consumptions of the software component from the future condition of the software component according to the component model of the software component.

### Claim 4 (depends on 3)

The method according to claim 3, wherein the method comprises:
reading, by the computing system for each of the plurality of software components, the component model from a manifest of the software component.

### Claim 5 (depends on 1)

The method according to claim 1, wherein the method comprises:
measuring, by the computing system for each of the plurality of software components, corresponding present consumptions of the computing resources each indicative of a present value of the consumption of the corresponding computing resource by the software component;
calculating, by the computing system for each of the plurality of software components, corresponding estimation errors of the computing resources, the estimation error of each of the computing resources being estimated according to the present consumption and a previous future consumption consisting of the future consumption being previously estimated; and
controlling, by the computing system, the allocation of the computing resources to the plurality of software components further according to the corresponding estimation errors.

### Claim 6 (depends on 5)

The method according to claim 5, wherein the method comprises:
refining, by the computing system for each of the plurality of software components, the future consumption of each of the computing resources according to the estimation error of the computing resource.

### Claim 7 (depends on 6)

The method according to claim 6, wherein the method comprises:
estimating, by the computing system for each of the plurality of software components, the future consumptions of the computing resources comprising corresponding baseline future consumptions, minimum future consumptions and maximum future consumptions, the baseline future consumption, the minimum future consumption and the maximum future consumption of each of the computing resources being indicative of a baseline, a minimum and a maximum, respectively, of the future value of the consumption of the corresponding computing resource;
calculating, by the computing system for each of the plurality of software components, the estimation errors comprising corresponding consumption errors, the consumption error of each of the computing resources being calculated according to a difference between the present consumption of the computing resource and a previous future consumption of the computing resource consisting of the baseline future consumption of the computing resource being previously estimated; and
refining, by the computing system for each of the plurality of software components, the minimum future consumption and the maximum future consumption of each of the computing resources according to the baseline future consumption and the consumption error of the computing resource.

### Claim 8 (depends on 7)

The method according to claim 7, wherein the method comprises:
setting, by the computing system for each of the plurality of software components, the minimum future consumption of each of the computing resources to a minimum between the minimum future consumption and the baseline future consumption minus the consumption error of the computing resource; and
setting, by the computing system for each of the plurality of software components, the maximum future consumption of each of the computing resources to a maximum between the maximum future consumption and the baseline future consumption plus the consumption error of the computing resource.

### Claim 9 (depends on 5)

The method according to claim 5, wherein the method comprises:
triggering, by the computing system for each of the plurality of software component, the controlling of the allocation of each of the computing resources according to a comparison of the future consumption of the computing resource with one or more corresponding verification thresholds based on the estimation error of the computing resource.

### Claim 10 (depends on 9)

The method according to claim 9, wherein the method comprises:
monitoring, by the computing system for each of the plurality of software components, the present condition of the software component comprising corresponding present metrics of one or more workload metrics of the software component each indicative of a present value of the corresponding workload metric;
estimating, by the computing system for each of the plurality of software components, corresponding future metrics for the workload metrics each indicative of a future value of the corresponding workload metric of the software component, the future metric of each of the workload metrics of the software component being estimated from the present metrics of the workload metric of the software components according to said at least one estimation model;
estimating, by the computing system for each of the plurality of software components, for each of the computing resources corresponding partial future consumptions for the workload metrics, the partial future consumption of each of the workload metrics being estimated from a previous future metric of the workload metric, consisting of the future metric of the workload metric being previously estimated, and the present metrics of any other workload metrics according to said at least one estimation model;
calculating, by the computing system for each of the plurality of software components, the estimation error of each of the computing resources comprising corresponding metric errors for the workload metrics, the metric error of each of the workload metric being calculated according to a difference between the present consumption of the computing resource and the partial future consumption of the workload metric for the computing resource; and
setting, by the computing system for each of the software component, the verification thresholds of each of computing resources according to the metric errors of the computing resource.

### Claim 11 (depends on 10)

The method according to claim 10, wherein the method comprises:
estimating, by the computing system for each of the plurality of software components, the future consumptions of the computing resources comprising corresponding minimum future consumptions and maximum future consumptions, the minimum future consumption and the maximum future consumption of each of the computing resources being indicative of a minimum and a maximum, respectively, of the future value of the consumption of the corresponding computing resource; and
triggering, by the computing system for each of the plurality of software component, the controlling of the allocation of each of the computing resources in response to the minimum future consumption of the computing resource being lower than a low one of the verification thresholds and the maximum future consumption of the computing resource being higher than a high one of the verification thresholds higher than the low verification threshold.

### Claim 12 (depends on 11)

The method according to claim 11, wherein the method comprises:
setting, by the computing system for each of the plurality of software component, the low verification threshold and the high verification threshold of each of the computing resources to a sum of workload metrics of the computing resource multiplied by a low scaling factor and a high scaling factor, respectively, the high scaling factor being higher than the low scaling factor.

### Claim 13 (depends on 1)

The method according to claim 1, wherein the software application is a cloud software application and the plurality of software components implement corresponding services.

### Claim 14

A computer program product for resource management of a software application comprising a plurality of software components interacting with each other, the computer program product comprising:
one or more computer readable tangible storage medium and program instructions stored on at least one of the one or more tangible storage medium, the program instructions executable by a computing system to cause the computing system to perform a method comprising:
monitoring corresponding present conditions of the plurality of software components;
estimating, for each of the plurality of software components, corresponding future consumptions of one or more computing resources each indicative of a future value of a consumption of the corresponding computing resource by the software component, the future consumptions of the software component being estimated from the present conditions of the software components according to at least one estimation model; and
controlling an allocation of the computing resources to the plurality of software components according to the corresponding future consumptions.

### Claim 15

A system for resource management of a software application comprising a plurality of software components interacting with each other, wherein the system comprises:
a monitor for monitoring corresponding present conditions of the software components;
an estimator for estimating, for each of the software components, corresponding future consumptions of one or more computing resources each indicative of a future value of a consumption of the corresponding computing resource by the software component, the future consumptions of the software component being estimated from the present conditions of the software components according to at least one estimation model; and
a controller for controlling an allocation of the computing resources to the software components according to the corresponding future consumptions.
