# US 2022/0035667 A1 · Resource availability-based workflow execution timing determination

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

- Number: US 2022/0035667 A1
- Document ID: US20220035667A1
- Status: Published application
- Kind: A1
- Application number: 17/505,367
- Filed: 2021-10-19
- Published: 2022-02-03
- Earliest claimed: 2019-05-29
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Yannick Saillet; Namit Kabra
- Web page: https://patentgrep.com/patent/US20220035667A1

## CPC

- G06F 9/4887: {involving deadlines, e.g. rate based, periodic}
- G06F 9/505: {considering the load}
- G06F 9/4881: {Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues}
- G06F 9/5027: {the resource being a machine, e.g. CPUs, Servers, Terminals}
- G06N 20/00: Machine learning
- G06F 2209/483: Multiproc
- G06F 2209/5019: Workload prediction
- G06F 2209/508: Monitor
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

According to a computer-implemented method, an available amount of each of multiple computing resources is determined by machine logic over a period of time at a computing device. The machine logic also determines an expected usage of each computing resource to execute each workflow in a queue. The machine logic also determines a time of execution of each workflow in the queue based on the available amount of each of the multiple computing resources over time and the expected usage of each computing resource to execute each workflow in the queue.

## Claims

### Claim 1

A computer-implemented method, comprising:
scheduling, by machine logic, multiple workflows in a queue for execution over a period of time by a data processing system;
determining, by the machine logic, an available amount of multiple computing resources of the data processing system to execute the multiple workflows over the period of time;
determining, by the machine logic, an expected usage of each computing resource to execute the multiple workflows in the queue; and
rescheduling, by the machine logic, the multiple workflows in the queue based on an execution order that maximizes usage of the multiple computing resources.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, wherein scheduling the multiple workflows in the queue comprises sequentially placing the multiple workflows in the queue in an order that the multiple workflows are received.

### Claim 3 (depends on 1)

The computer-implemented method of claim 1, wherein scheduling the multiple workflows in the queue comprises placing the multiple workflows in the queue based on a priority of the multiple workflows.

### Claim 4 (depends on 1)

The computer-implemented method of claim 1, wherein determining the expected usage of each computing resource to execute the multiple workflows in the queue comprises mapping the multiple workflows to previous workflows that have been performed based on characteristics of the multiple workflows.

### Claim 5 (depends on 4)

The computer-implemented method of claim 4, further comprising storing historical information of an amount of computing resources used over time for the previous workflows.

### Claim 6 (depends on 1)

The computer-implemented method of claim 1, wherein rescheduling the multiple workflows comprises ordering the multiple workflows in the queue based on an expected average computing resource usage of the multiple workflows in the queue.

### Claim 7 (depends on 6)

The computer-implemented method of claim 6, wherein the multiple workflows are ordered in the queue in descending order of expected average computing resource usage, wherein a workflow with a highest expected average computing resource usage is executed first.

### Claim 8

A system, comprising:
a scheduler to schedule multiple workflows in a queue for execution over a period of time by a data processing system;
a resource analyzer to determine an available amount of multiple computing resources of the data processing system to execute the multiple workflows over the period of time; and
a workflow analyzer to determine an expected usage of each computing resource to execute the multiple workflows in the queue; wherein
the scheduler to reschedule the multiple workflows in the queue based on an execution order that avoids overcommitment of the multiple computing resources.

### Claim 9 (depends on 8)

The system of claim 8, wherein the scheduler is to:
sequentially analyze each of the multiple workflows in the queue to determine whether a given workflow would overcommit any of the multiple computing resources at a scheduled time of execution; and
move the given workflow that would overcommit any of the multiple computing resources down in the queue.

### Claim 10 (depends on 8)

The system of claim 8, wherein the scheduler is to:
determine which of the multiple workflows can be executed at a particular point in time without overcommitting the multiple computing resources; and
determine, for workflows that would overcommit the multiple computing resources if executed at the particular point in time, when each of the workflows can be executed without overcommitting the multiple computing resources.

### Claim 11 (depends on 8)

The system of claim 8, wherein the scheduler is to:
determine that expected computing resource usage for a given workflow is above a threshold level; and
reschedule the given workflow in the queue in response to determining that the expected computing resource usage for the given workflow would not overcommit the multiple computing resources over the time period.

### Claim 12 (depends on 11)

The system of claim 11, wherein determining that the expected computing resource usage for the given workflow would not overcommit the multiple computing resources over the time period comprises determining that the expected computing resource usage over the time period is less than a maximum resource utilization limit.

### Claim 13 (depends on 11)

The system of claim 11, wherein the threshold level is a user-specified parameter.

### Claim 14

A computer-implemented method, comprising:
scheduling, by machine logic, a number of workflows in a scheduling system for execution by a data processing system;
creating a superimposed time-based graph of expected computing resource usage by the workflows to decide which workflow to execute at which point of time; and
rescheduling the workflows based on the superimposed time-based graph to form a queue of workflows wherein the multiple computing resource usage is maximized.

### Claim 15 (depends on 14)

The computer-implemented method of claim 14, wherein creating the superimposed time-based graph comprises depicting different computing resource usage levels over time on different time-based graphs.

### Claim 16 (depends on 14)

The computer-implemented method of claim 14, wherein creating the superimposed time-based graph comprises depicting different expected computing resource usage levels over time on a same graph.

### Claim 17 (depends on 14)

The computer-implemented method of claim 14, wherein the superimposed time-based graph superimposes expected usage of each computing resource to execute each workflow in the queue.

### Claim 18 (depends on 14)

The computer-implemented method of claim 14, wherein the superimposed time-based graph indicates a current availability of a computing resource and an expected availability of the computing resource if the queued workflow was executed.

### Claim 19 (depends on 14)

The computer-implemented method of claim 14, further comprising:
generating multiple superimposed expected computing resource usages of different workflows on the superimposed time-based graph; and
determining a time of execution of each workflow based on a comparison of the multiple superimposed expected computing resource usages of the different workflows on the superimposed time-based graph.

### Claim 20 (depends on 19)

The computer-implemented method of claim 19, wherein determining the time of execution of each workflow comprises selecting a given workflow with a highest expected computing resource usage as compared to other workflows in the superimposed time-based graph.
