Dynamic Microservices Allocation via Infrastructure Processing Unit
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current orchestrators in disaggregated computing environments cannot dynamically compose platforms of disaggregated hardware resources per customer requirement, nor can they provision knowledge of available resources, allocate them efficiently, or set up communications amongst resources, leading to inefficiencies in resource utilization and microservice deployment.
Innovation Solution
An infrastructure processing unit (IPU) is used to dynamically compose and manage disaggregated data center resources, enabling a cloud service provider to construct platforms on demand based on available resources, and an orchestration controller schedules microservices on suitable worker nodes based on resource requirements, allowing for flexible and efficient resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If current orchestrators are used in disaggregated computing environments, then microservice deployment can be performed, but dynamic platform composition and efficient resource allocation cannot be achieved
Solution Approach 1:
The patent implements dynamic resource allocation by allowing the platform to compose and reconfigure disaggregated hardware resources in real-time based on workload demands. The resource manager dynamically assigns CPU, memory, storage, and network resources to microservices, enabling the system to adapt its platform composition rather than relying on static pre-provisioned configurations.
Solution Approach 2:
The patent creates a universal resource management layer that can allocate multiple types of hardware resources (CPU, memory, storage, network) through a single orchestration mechanism. This multi-functional approach allows the same infrastructure to serve diverse microservice workloads with different resource requirements, achieving both adaptability and efficiency.
2Productivity
If static resource provisioning is used, then system simplicity is maintained, but resource allocation efficiency deteriorates
Solution Approach 1:
The patent introduces an intermediary resource manager layer between the physical hardware infrastructure and the microservices. This mediator handles the complexity of dynamic resource allocation, translating high-level microservice requirements into low-level resource assignments. The intermediary absorbs the orchestration complexity while presenting a simplified interface to both hardware and applications.
Solution Approach 2:
The resource allocation system operates autonomously by automatically discovering available hardware resources, matching them to microservice requirements, and performing assignments without manual intervention. The system self-manages the complexity of resource composition and reconfiguration, reducing the operational burden while maintaining high allocation efficiency.
3Adaptability or versatility
If disaggregated computing is implemented, then scalability and vendor flexibility improve, but resource composition and communication setup become complex
Solution Approach 1:
The patent segments the resource management functionality into distinct modular components: resource discovery, resource characterization, resource matching, and resource assignment. Each component handles a specific aspect of the composition process, making the overall complex system manageable through clear separation of concerns. This segmentation allows independent optimization of each module while maintaining vendor flexibility.
Data Source
AI summary
A computing platform comprising a plurality of disaggregated data center resources and an infrastructure processing unit (IPU), communicatively coupled to the plurality of resources, to compose a platform of the plurality of disaggregated data center resources for allocation of micro service s cluster.