Logical Partition Load Balancer with Preliminary Migration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In logical partitioning data processing systems, managing and balancing load across multiple managed systems is challenging, especially when critical partitions require additional resources, as existing methods struggle to dynamically reallocate resources without disrupting operations or leaving unused resources on source systems after partition mobility.

Innovation Solution

The method combines Live Partition Mobility (LPM) with dynamic logical partitioning (DLPAR) to migrate and reallocate resources across managed systems, ensuring that critical partitions receive necessary resources while optimizing resource utilization on both source and target systems.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If Live Partition Mobility is used to migrate critical partitions to other managed systems, then resource availability for critical partitions is improved, but resource utilization efficiency on source systems deteriorates due to leftover unused resources

Engineering Contradiction:
Improveresource availability for critical partitionsVSAvoidresource utilization efficiency
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The system performs preliminary actions by migrating non-critical partitions to target managed systems before the critical partition migration is complete. This preliminary migration of other partitions ensures that when the critical partition is migrated, the target system already has resources allocated and ready, preventing resource waste and optimizing utilization efficiency while maintaining reliability.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If dynamic reallocation of resources is performed to satisfy critical partition requirements, then partition performance is improved, but system complexity increases due to coordination between multiple managed systems

Engineering Contradiction:
Improvepartition performanceVSAvoidsystem coordination complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system introduces an intermediary mechanism that coordinates between source and target managed systems during resource reallocation. This intermediary manages the complexity by orchestrating the migration process, ensuring proper resource allocation, and maintaining system stability while improving partition performance through dynamic reallocation.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Adaptability or versatility

If partitions are migrated between managed systems, then resource distribution is improved, but operational disruption may occur during the migration process

Engineering Contradiction:
Improveresource distribution flexibilityVSAvoidoperational stability
Core Design Contradiction:
Adaptability or versatilityVSStability of the object's composition

Solution Approach 1:

The system performs preliminary migration of non-critical partitions to the target managed system before migrating the critical partition. This preliminary action prepares the target system with necessary resources and configuration, allowing the critical partition migration to proceed without operational disruption and maintaining stability while achieving flexible resource distribution.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS8661448B2Logical partition load manager and balancer
Publication Date: 2014.02.25 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US8661448B2 patent drawing
  • US8661448B2 patent drawing
  • US8661448B2 patent drawing

AI summary

A mechanism is provided in a data processing system for managing and balancing load in multiple managed systems in a logical partitioning data processing system. Responsive to a critical logical partition requiring additional resources, the mechanism determines whether one or more managed systems have available resources to satisfy resource requirements of the critical partition. The mechanism performs at least one partition migration operation to move at least one logical partition between managed systems responsive to determining that one or more managed systems have available resources to satisfy resource requirements of the critical partition. The mechanism performs at least one dynamic logical partitioning operation to allocate resources to at least one of the one or more critical logical partitions responsive to performing the at least one partition migration operation.