Universal Placer for Multi-Cloud Workload Deployment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multicloud environments increase management complexity and risk vendor lock-in due to interactions with multiple cloud vendors, making it challenging to efficiently deploy workloads and manage resources based on workload requirements.
Innovation Solution
A computer-implemented method using open-source container-orchestration tools like Kubernetes to create a universal public cloud, allowing tenants to specify workload requirements and receive cost estimates without interacting directly with underlying cloud providers, thereby determining optimal cluster placement across multiple cloud providers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If direct interaction with multiple cloud vendors is used for workload deployment, then workload deployment flexibility is improved, but management complexity increases
Solution Approach 1:
The patent introduces a universal placer as an intermediary system that sits between the workload deployment request and multiple cloud providers. This mediator receives workload requirements, evaluates multiple cloud options, and handles the deployment process, thereby reducing management complexity while maintaining deployment flexibility across different cloud vendors.
Solution Approach 2:
The universal placer provides a unified interface that can deploy workloads to any supported cloud provider using standardized criteria. This multi-functional system handles diverse cloud environments through a single consistent mechanism, simplifying management while preserving adaptability across different cloud platforms.
2Ease of operation
If direct interaction with cloud vendors is required for deployment, then deployment control is improved, but vendor lock-in risk increases
Solution Approach 1:
The universal placer acts as an intermediary that decouples the deployment control from direct vendor interactions. It maintains control over deployment decisions while using standardized interfaces to communicate with cloud providers, thereby reducing vendor lock-in risk while preserving deployment control.
Solution Approach 2:
The system segments the deployment process into independent evaluation criteria and standardized interface layers. This segmentation allows the universal placer to maintain control through high-level decision-making while using standardized, vendor-agnostic interfaces at the execution level, reducing dependency on any single vendor.
3Productivity
If multiple cloud providers are evaluated for optimal placement, then workload placement optimization is improved, but evaluation complexity increases
Solution Approach 1:
The universal placer transforms complex cloud provider evaluations into standardized parameter comparisons. By converting diverse cloud offerings into uniform evaluation criteria (such as cost, performance, availability), the system achieves optimized workload placement while reducing evaluation complexity through parameter standardization.
Solution Approach 2:
The system implements feedback mechanisms that continuously monitor workload performance and cloud provider metrics. This feedback enables automated optimization of workload placement by comparing actual performance against evaluation criteria, reducing manual evaluation complexity while improving placement optimization through data-driven decisions.
Data Source
AI summary
In an approach, a processor receives a request to deploy a workload in a container environment, where: the container environment comprises a plurality of external providers running container environment clusters; and the request (i) includes one or more requirements of the workload and (ii) does not specify a particular external provider of the plurality of external providers. A processor determines a cluster, from the plurality of external providers running the container environment clusters, that meets the one or more requirements of the workload. A processor deploys the workload on the determined cluster.