Intelligent Workload Scheduling for Batch and Transaction Systems

Solution Overview

Problem

Businesses face challenges in managing transaction peak times due to interference between online transaction workloads and batch job workloads, leading to resource conflicts and delayed transactions, as existing solutions either reserve resources inefficiently or cause delays by pausing or terminating batch jobs.

Innovation Solution

Intelligent workload scheduling that classifies batch jobs by resource requests and dependencies, schedules them to balance resource usage, monitors system resources, and reduces batch job resource quotas to ensure transactions complete within a set time threshold, thereby avoiding peak times.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If batch jobs are scheduled during batch windows when CPU usage is low, then batch workloads can require high usage of the CPU without interfering with operational processes, but batch workloads may cause transaction response time to exceed the pre-set threshold during peak times

Engineering Contradiction:

Improvebatch processing efficiencyVSAvoidtransaction completion time guarantee

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent implements dynamic resource allocation where the workload automation system continuously monitors system resources and transaction queue depths, adjusting batch job scheduling and resource allocation in real-time based on current system conditions, allowing the system to adapt between peak and off-peak periods

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system employs feedback mechanisms by monitoring transaction response times, queue depths, and resource usage metrics, then using this information to dynamically adjust batch job scheduling decisions, ensuring transaction SLAs are met while optimizing batch processing efficiency

Inventive Principle:

Principle #23Feedback

2Productivity

If resources are reserved for batch jobs, then batch processing can proceed without resource conflicts, but resources may not be available for transaction workloads during peak times

Engineering Contradiction:

Improvebatch processing throughputVSAvoidtransaction processing speed

Core Design Contradiction:

ProductivityVSSpeed

Solution Approach 1:

The patent implements dynamic resource allocation where the workload automation system continuously monitors system resources and transaction queue depths, adjusting batch job scheduling and resource allocation in real-time based on current system conditions, allowing the system to adapt between peak and off-peak periods

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes resource allocation parameters dynamically based on system state, adjusting CPU and memory allocation for batch jobs versus transaction processing based on monitored metrics such as transaction response time, queue depth, and resource utilization levels

Inventive Principle:

Principle #35Parameter changes

3Reliability

If batch jobs are paused or terminated to release resources for transactions, then transaction response time can be improved, but batch job completion is delayed

Engineering Contradiction:

Improvetransaction completion time guaranteeVSAvoidbatch job completion delay

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent implements dynamic resource allocation where the workload automation system continuously monitors system resources and transaction queue depths, adjusting batch job scheduling and resource allocation in real-time based on current system conditions, allowing the system to adapt between peak and off-peak periods

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system implements periodic monitoring and adjustment of batch job execution, pausing and resuming batch jobs in cycles based on transaction workload conditions, allowing batch processing to make progress during low-utilization periods while ensuring transaction SLAs are met during peak periods

Inventive Principle:

Principle #19Periodic action

Data Source

PatentUS20240419505A1Intelligent workload scheduling

Publication Date: 2024.12.19 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

In an approach for intelligent workload scheduling, a processor groups a plurality of batch jobs based on workload resource requests and dependencies of each batch job resulting in a plurality of groups. A processor schedules the plurality of batch jobs based on the plurality of groups. A processor monitors workload resource usage of system for running the plurality of batch jobs and a plurality of transaction workloads. A processor identifies one or more scheduled transaction workloads will not be able to be completed in under a preset time threshold. A processor reduces a resource quota of one or more batch jobs of the plurality of batch jobs based on type of resource that is needed for the one or more scheduled transaction workloads.