Processing Unit Subtype Configuration for Dynamic Resource Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computer systems face inefficiencies due to static allocation of processing resources, leading to reduced performance as new workloads are installed, as existing technologies do not dynamically adjust resource allocation based on workload demands.
Innovation Solution
The solution involves using processing unit subtype identifiers to dynamically allocate resources by associating software image types with specific processing unit allocations, enabling efficient pooling and sharing of resources across logical partitions, and managing resource utilization through license records.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If processing resources are statically allocated to logical partitions, then reliability is improved, but productivity deteriorates due to blocked shared resources
Solution Approach 1:
The patent implements dynamic resource allocation by allowing logical partitions to access shared processing resources on-demand rather than through static assignment. The resource management system monitors workload demands and allocates processing units dynamically, enabling the system to adapt resource distribution in real-time based on actual needs, thus improving both reliability and productivity
Solution Approach 2:
The patent creates a universal resource pool where processing units can serve multiple logical partitions simultaneously based on demand. Instead of dedicating specific resources to specific partitions, the system makes processing resources universally accessible to any logical partition that requires them, eliminating resource blocking and improving overall system productivity
2Productivity
If more processing resources are assigned to logical partitions, then productivity is improved, but loss of substance worsens due to inefficient resource utilization
Solution Approach 1:
The patent enables logical partitions to self-serve from the shared resource pool based on their own workload demands. Each logical partition can access processing resources independently when needed, eliminating the need for over-provisioning and ensuring that resources are utilized efficiently without remaining idle or unused
Solution Approach 2:
The resource management system implements feedback mechanisms to monitor the actual utilization of processing resources by logical partitions. Based on this feedback, the system dynamically adjusts resource allocation to match actual demand, preventing both resource shortages and inefficient over-allocation, thus reducing loss of unused processing resources
3Productivity
If processing resources are shared across logical partitions, then productivity is improved, but device complexity increases due to resource management overhead
Solution Approach 1:
The patent introduces an intermediary resource management system that sits between the physical processing resources and logical partitions. This intermediary layer abstracts the complexity of dynamic resource allocation, presenting a simplified interface to logical partitions while handling the complex tasks of monitoring, allocating, and managing shared resources efficiently
Data Source
AI summary
An aspect includes receiving a software image file set and a capacity requirement at a software image distribution system. A software image is generated based on the software image file set. A license record is generated based on the capacity requirement. The software image and the license record are provided to an external interface of the software image distribution system. An installation action is triggered by the software image distribution system on to a machine based on a request of an ordering system.