Multi-Cloud Serverless Workflow Scheduling With PSO Cost Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for scheduling complex task workflows in a multi-cloud environment using serverless architectures are inefficient in minimizing overall deployment costs, as they often rely on virtual machines and fail to optimize the use of scalable and cost-effective serverless platforms, considering the unique cost models and performance of different cloud service providers.
Innovation Solution
A processor-implemented method using particle swarm optimization (PSO) to determine optimal mappings of tasks with serverless computing instances and storage services in a multi-cloud environment, minimizing a multi-objective optimization function that balances makespan and execution cost, by considering compute requirements, I/O requirements, and bandwidth measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional algorithms use virtual machines for scheduling complex task workflows in multi-cloud environments, then infrastructure availability is improved, but deployment cost and scalability are worsened
Solution Approach 1:
The patent changes the fundamental parameter of compute resource type from virtual machines to serverless computing instances. This parameter change enables cost optimization through pay-per-execution pricing models while maintaining reliability through automatic scaling and stateless architecture, directly resolving the contradiction between infrastructure availability and deployment cost
Solution Approach 2:
The patent replaces the mechanical VM-based scheduling system with a serverless computing system. This substitution eliminates the need for manual VM provisioning and management, reducing operational complexity and cost while improving scalability through automatic resource allocation based on task requirements
2Ease of manufacture
If serverless platforms are used for scheduling complex task workflows, then cost-effectiveness and scalability are improved, but scheduling optimization efficiency is worsened
Solution Approach 1:
The patent implements feedback mechanisms in the PSO algorithm to continuously monitor and adjust scheduling decisions based on actual task execution metrics. This feedback loop enables the system to optimize cost-effectiveness while maintaining high scheduling efficiency by learning from previous deployments and adapting to changing workload patterns
Solution Approach 2:
The patent introduces dynamic scheduling capabilities that allow the system to adapt resource allocation in real-time based on task requirements and cloud provider conditions. This dynamic approach enables cost optimization through flexible resource usage while maintaining scheduling efficiency through automated decision-making algorithms
3Adaptability or versatility
If heterogeneous serverless computing instances are deployed across multiple cloud providers, then service versatility and cost optimization are improved, but system complexity and mapping difficulty are worsened
Solution Approach 1:
The patent segments the complex scheduling problem into independent modules: task analysis, cloud provider evaluation, and optimization mapping. This segmentation allows the system to handle heterogeneous serverless instances across multiple cloud providers by processing each component separately, reducing overall system complexity while maintaining service versatility
Solution Approach 2:
The patent creates a universal scheduling framework that can handle diverse serverless computing instances from different cloud providers through a unified optimization approach. This multi-functional system abstracts the complexities of heterogeneous architectures behind a common interface, enabling versatile deployment while managing system complexity through standardization
4Productivity
If task mapping is optimized considering multiple objectives (cost, performance, I/O requirements), then deployment optimization is improved, but computational time and processing complexity are worsened
Solution Approach 1:
The patent applies partial optimization by focusing the PSO algorithm on the most critical optimization objectives (cost and makespan) while using heuristics for secondary considerations. This partial action approach achieves significant deployment optimization benefits while reducing computational time by avoiding exhaustive optimization of all parameters simultaneously
Data Source
AI summary
The disclosure relates generally to methods and systems for multi-objective workflow scheduling to serverless architecture in a multi-cloud environment. Generating an optimal mapping scheme for heterogeneous tasks of the complex task workflow in the multi-cloud environment is always a challenge. The present disclosure make use of the serverless platforms in conjunction with the storage services for the optimal mapping of the task workflows to the multi-cloud environment using the particle swarm optimization (PSO) algorithm. In the present disclosure, each of the tasks of the application are characterized to determine one or more compute requirements, and one or more input/output (I/O) requirements. Furthermore, a compute capacity of each of the plurality of serverless computing instances, and bandwidth measurements are determined. Then, the optimized task workflow is generated, using the PSO technique by minimizing a multi-objective optimization function.