Microservice Deployment Scheduler With RL Workload Orchestration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing the deployment lifecycles of microservices across a vast number of computing nodes in distributed computing environments, such as edge computing, is challenging due to the mismatch between computing node properties and microservice metadata constraints.
Innovation Solution
A reinforcement learning-based dynamic workload orchestration module in a microservice deployment scheduler that receives microservice constraints and node-specific properties to efficiently allocate and orchestrate computing resources, utilizing a trained machine learning module to classify and optimize node-specific properties for effective deployment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional deployment methods are used to manage microservice deployment across vast number of computing nodes, then deployment can be achieved, but resource utilization efficiency deteriorates and scalability is limited
Solution Approach 1:
The patent implements dynamic workload orchestration using reinforcement learning that continuously adapts to changing system states. The orchestration module dynamically adjusts microservice deployment decisions based on real-time node properties, constraints, and system performance, enabling the system to scale efficiently across vast numbers of computing nodes while maintaining optimal resource utilization
Solution Approach 2:
The system changes deployment parameters dynamically by using reinforcement learning to optimize assignment decisions. The orchestration module adjusts assignment parameters based on learned patterns from system state observations, transforming static deployment configurations into adaptive, parameter-optimized deployments that improve both efficiency and scalability
2Ease of operation
If microservice constraints and node properties are not properly matched, then deployment simplicity is maintained, but resource utilization efficiency deteriorates
Solution Approach 1:
The reinforcement learning-based orchestration module autonomously performs the complex task of matching microservice constraints with node properties without requiring manual intervention. The system self-optimizes deployment assignments by learning from system states and constraints, maintaining deployment simplicity while achieving high resource utilization efficiency through automated, intelligent decision-making
Solution Approach 2:
The system implements feedback mechanisms where the orchestration module continuously observes system states, evaluates deployment outcomes, and uses reinforcement learning to refine future assignment decisions. This feedback loop ensures that microservice constraints are properly matched with node properties, improving resource utilization while keeping the deployment process simple and automated
3Productivity
If reinforcement learning-based dynamic workload orchestration is implemented, then resource utilization and scalability are improved, but system complexity increases
Solution Approach 1:
The patent introduces a reinforcement learning-based orchestration module as an intermediary between microservice deployment requests and computing node assignments. This intermediary handles the complexity of matching constraints and optimizing resource allocation, shielding users from system complexity while improving resource utilization and scalability through intelligent, automated decision-making
4Adaptability or versatility
If dynamic workload orchestration is used, then adaptability to different deployment scenarios is improved, but computational overhead increases
Solution Approach 1:
The reinforcement learning model performs preliminary learning and pattern recognition during system operation, building knowledge about optimal deployment strategies in advance. This preliminary action enables the orchestration module to make rapid, adaptive decisions about microservice assignments without requiring heavy computational overhead during actual deployment operations, balancing adaptability with computational efficiency
Data Source
AI summary
A computer implemented method of deployment and resource allocation of microservices of a distributed computing environment is disclosed. The distributed computing environment includes a microservice deployment scheduler and one or more computing nodes. The microservice deployment scheduler includes a reinforcement learning based dynamic workload orchestration module. The method includes receiving microservice constraints descriptive of a microservice computing task by the microservice deployment scheduler. The method further includes receiving node specific properties from the one or more computing nodes by the microservice deployment scheduler. The node specific properties are descriptive of a computing capacity and/or computing capabilities of the one or more computing nodes. The method further includes orchestrating operation of the one or more computing nodes by the microservice deployment scheduler by inputting the microservice constraints and the node specific properties into the reinforcement learning based dynamic workload orchestration module.