Federated Controller Workflow Decomposition for Device Ecosystems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In complex device ecosystems, determining the optimal deployment of workflows across diverse devices and platforms becomes increasingly difficult as the ecosystem grows in size and complexity, making it challenging to inventory and characterize connectivity to support complex workflows and meet service level objectives (SLOs).
Innovation Solution
A method involving a federated controller system that decomposes workflows into portions, assigns them to appropriate platform controllers based on capabilities and constraints, and initiates data management services to optimize deployment and execution within the ecosystem, utilizing a graph or database to map ecosystem information and determine the best placement for workflow portions to minimize data transfer and ensure service objectives are met.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If workflows are deployed in complex device ecosystems with more devices and data, then the system's capability to process and operate data is improved, but the difficulty of determining where to deploy workflows and how to efficiently execute them increases
Solution Approach 1:
The workflow is decomposed into multiple workflow portions that can be independently assigned to different platform controllers. This segmentation allows the complex deployment problem to be broken down into smaller, more manageable sub-problems, where each portion can be optimally placed based on local ecosystem characteristics and capabilities.
Solution Approach 2:
A service controller acts as an intermediary between the workflow deployment request and the platform controllers. The service controller receives the workflow, decomposes it into portions, and assigns them to appropriate platform controllers based on ecosystem capabilities and constraints. This intermediary manages the complexity of deployment determination centrally while enabling distributed execution.
2Quantity of substance
If the ecosystem size and complexity increase with more data and devices, then the system's data processing capacity is improved, but the difficulty of inventorying and characterizing connectivity to support workflows increases
Solution Approach 1:
Each platform controller maintains local knowledge of its ecosystem domain, including inventory of devices, data locations, and connectivity characteristics. This local quality approach allows the system to characterize connectivity and resources without requiring a centralized view of the entire complex ecosystem, reducing the difficulty of inventory and characterization.
Solution Approach 2:
Platform controllers provide feedback to the service controller about their capabilities, available resources, and execution status. This feedback mechanism enables the service controller to make informed deployment decisions based on current ecosystem state, effectively managing the complexity of inventory and connectivity characterization through distributed information gathering.
3Productivity
If workflows are decomposed and assigned to multiple platform controllers, then the efficiency of workflow execution is improved, but the complexity of managing data management services across controllers increases
Solution Approach 1:
The service controller provides universal data management service coordination across multiple platform controllers. It handles data location determination, data transfer coordination, and service assignment uniformly regardless of which platform controllers are involved. This universal approach manages the complexity of coordinating data management services across distributed controllers.
Data Source
AI summary
Techniques described herein relate to a method for deploying workflows with data management services. The method may include receiving, at a service controller of a federated controller, a request to deploy a workflow in a device ecosystem; decomposing the workflow into a plurality of workflow portions; assigning, based on identifying a first platform controller, a first workflow portion to the first platform controller; assigning, based on identifying a second platform controller, a second workflow portion to the second platform controller; assigning, based on the plurality of workflow portions and a capabilities and constraints data construct, data management services to the first platform controller and the second platform controller; and initiating performance of the first workflow portion by the first platform controller and the second workflow portion by the second platform controller, in which performing the first workflow portion and the second workflow portion includes performing the data management services.