SDI Partition Manager for Composite Node Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data center architectures face challenges in correctly allocating and managing resources, particularly in software-defined infrastructure (SDI) systems, where existing methods struggle to efficiently allocate resources to meet application demands, leading to suboptimal performance and throughput.
Innovation Solution
The implementation of static or dynamic partitions, which can be user-defined or determined using machine learning techniques, to manage resources within data centers, ensuring optimal allocation and quality of service (QoS) for applications, by pre-populating partitions with composite nodes composed of disaggregated resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If resources are allocated dynamically in SDI systems, then resource flexibility and adaptability improve, but node creation and boot time increase
Solution Approach 1:
The patent applies preliminary action by pre-composing node partitions before they are needed. The system pre-assembles groups of disaggregated resources (compute, storage, memory, fabric) into ready-to-use node partitions and stores them in a partition table. When an application needs a node, the pre-composed partition is immediately allocated without requiring real-time resource assembly, thus reducing node creation and boot time while maintaining resource flexibility.
2Manufacturing precision
If composite nodes are created on-demand, then resource allocation precision improves, but system complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the data center resources into distinct, manageable partitions. Each partition is composed of specific disaggregated resources (compute nodes, storage nodes, memory nodes, fabric nodes) that are pre-assembled into coherent groups. This segmentation allows the system to manage complex resource allocation by working with discrete, pre-defined partition units rather than individually managing each resource type, thus reducing operational complexity while maintaining allocation precision.
Solution Approach 2:
The patent introduces an intermediary layer (the partition manager and partition table) that mediates between resource requests and physical resources. This intermediary pre-composes node partitions and manages the mapping between applications and resources, simplifying the allocation process while ensuring precise resource allocation. The partition table acts as an intermediary data structure that stores pre-composed partition information, reducing the complexity of real-time resource management.
3Productivity
If disaggregated resources are used, then resource utilization efficiency improves, but QoS guarantee difficulty increases
Solution Approach 1:
The patent applies merging by combining disaggregated resources into unified node partitions that function as cohesive units. Each partition merges compute, storage, memory, and fabric resources into a integrated composite node that can be allocated as a single unit to an application. This merging ensures that all necessary resource types are available together within each partition, guaranteeing QoS requirements while maintaining the efficiency benefits of disaggregated resource pools.
Data Source
AI summary
Techniques for managing static and dynamic partitions in software-defined infrastructures (SDI) are described. An SDI manager component may include one or more processor circuits to access one or more resources. The SDI manager component may include a partition manager to create one or more partitions using the one or more resources, the one or more partitions each including a plurality of nodes of a similar resource type. The SDI manager may generate an update to a pre-composed partition table, stored within a non-transitory computer-readable storage medium, including the created one or more partitions, and receive a request from an orchestrator for a node. The SDI manager may select one of the created one or more partitions to the orchestrator based upon the pre-composed partition table, and identify the selected partition to the orchestrator. Other embodiments are described and claimed.