Logical Partition Allocation Across Mixed-Core Drawers

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing computing environments with multiple processor core types face challenges in efficiently allocating logical partitions across hardware drawers, leading to cache pollution and inconsistent performance due to varying hardware generations and configurations.

Innovation Solution

A method for allocating logical partitions that considers processor core types, allowing partitions to be placed on a single drawer when possible, and using bitmasks to distribute across multiple drawers while minimizing cross-drawer traffic and maintaining consistency, employing a modified fill and spill algorithm to optimize performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If logical partitions are allocated across multiple drawers to accommodate large partition sizes, then the partition capacity is improved, but cache pollution increases and performance consistency deteriorates

Engineering Contradiction:

Improvepartition capacityVSAvoidperformance consistency

Core Design Contradiction:

Quantity of substanceVSReliability

Solution Approach 1:

The system segments the allocation decision into two parts: first determines if a single drawer can accommodate the partition, and if not, then allocates across multiple drawers using bitmasks. This segmentation allows the system to prefer single-drawer allocations (avoiding cache pollution) while still supporting large partitions that require multiple drawers.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system applies different allocation strategies based on the specific characteristics of each partition and drawer configuration. By using bitmasks to represent processor chips and selectively allocating to specific drawers, the system tailors the allocation quality to local conditions, minimizing cache pollution while accommodating partition size requirements.

Inventive Principle:

Principle #3Local quality

2Reliability

If logical partitions are confined to a single drawer to minimize cache pollution, then performance consistency is improved, but the ability to accommodate large partitions deteriorates

Engineering Contradiction:

Improveperformance consistencyVSAvoidpartition capacity

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The system dynamically adjusts the allocation strategy based on partition size requirements. The hypervisor evaluates whether a single drawer can accommodate the partition, and only when necessary does it expand to multiple drawers. This dynamic approach allows the system to maintain performance consistency for smaller partitions while accommodating larger partitions when needed.

Inventive Principle:

Principle #15Dynamics

3Productivity

If the hypervisor uses complex allocation algorithms to optimize partition placement, then resource allocation efficiency is improved, but system complexity increases

Engineering Contradiction:

Improveresource allocation efficiencyVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system introduces bitmasks as an intermediary data structure to represent processor chip allocations. This intermediary simplifies the complex task of tracking and managing processor assignments across multiple drawers, making the allocation algorithm more manageable and efficient without significantly increasing overall system complexity.

Inventive Principle:

Principle #24Intermediary (Mediator)

4Speed

If the system allocates partitions without considering processor core types, then allocation speed is improved, but hardware compatibility and performance optimization deteriorate

Engineering Contradiction:

Improveallocation speedVSAvoidhardware compatibility

Core Design Contradiction:

SpeedVSReliability

Solution Approach 1:

The system performs preliminary analysis of processor core types and drawer configurations before allocating partitions. By pre-evaluating hardware characteristics and determining suitable drawer assignments in advance, the system maintains allocation speed while ensuring hardware compatibility and performance optimization. The bitmask representation enables efficient pre-computation of valid allocation options.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20250342067A1Allocating logical partitions in a computing environment with multiple processor core types

Publication Date: 2025.11.06 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A method includes receiving a request to create a logical partition. The method further includes determining whether the logical partition can be implemented on a single drawer of a plurality of drawers. The method further includes, responsive to determining that the logical partition can be implemented on a single drawer of the plurality of drawers, allocating the logical partition to one of the drawers of the plurality of drawers based on a container size that can fit the logical partition. The method further includes, responsive to determining that the logical partition cannot be implemented on a single drawer of the plurality of drawers, allocating the logical partition to at least two drawers of the plurality of drawers using bitmasks, wherein each of the bitmasks represents a processor chip of a plurality of processor chips, and wherein one bitmask is generated per drawer of the plurality of drawers.