Hybrid Cloud Workload Management via Dynamic Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Companies with private compute clouds face resource shortages during peak workload periods, leading to delayed product releases, and existing hybrid cloud solutions lack seamless task execution, intelligent resource scaling, and efficient data management between private and public clouds.
Innovation Solution
A hybrid cloud workload management system that automatically detects data dependencies, dynamically scales compute resources, and transparently shifts tasks from a private cloud to a public cloud when resources are insufficient, ensuring seamless execution and efficient resource utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If compute resources are expanded to handle peak workloads, then productivity is improved, but device complexity and cost increase
Solution Approach 1:
The system dynamically scales compute resources by creating additional virtual machine hosts in the public cloud when workload demands exceed private cloud capacity. The workload management system monitors resource utilization and automatically provisions or deprovisions hosts based on real-time workload conditions, enabling flexible adaptation without permanent infrastructure expansion.
Solution Approach 2:
The hybrid cloud architecture allows the same workload management system to manage workloads across both private and public cloud infrastructures. The system can execute tasks on either cloud environment based on resource availability, providing multi-functional capability that handles both peak and non-peak workloads through a unified management platform.
2Speed
If data is pre-synchronized to public cloud, then task execution speed is improved, but data storage cost and complexity increase
Solution Approach 1:
The system performs preliminary data synchronization by identifying and pre-copying only the specific data files required for upcoming tasks from the private cloud to the public cloud. The workload management system analyzes task requirements and proactively prepares necessary data in advance, ensuring fast task execution without maintaining continuous full data replication.
Solution Approach 2:
The system applies different data synchronization strategies to different data files based on their specific requirements. Only data files that are actually needed for task execution are synchronized to the public cloud, while other data remains in the private cloud. This selective approach optimizes the balance between execution speed and storage costs.
Data Source
AI summary
The present embodiments relate generally to workload management and more particularly to a hybrid cloud workload management system and methodology which can effectively manage the execution of tasks of the same workload on both private and public clouds. In embodiments, user tasks are seamlessly and transparently executed on a public cloud if the private cloud does not have the necessary resources available. These and other embodiments automatically detect data dependencies of user tasks and build lists of data attributes of user tasks, which are used to populate and synchronize data needed for tasks before they are executed on the public cloud. Additional or alternative embodiments include the ability to intelligently scale the compute resources in the public cloud so that appropriate number of hosts with the resources needed by the user tasks are dynamically created and also properly purged upon user task completion.