Linear Programming Resource Allocation for Thermal and Redundancy Constraints

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional resource allocation methods in virtualized data centers do not consider operational considerations such as heat dissipation, single points of failure, and resource compatibility, leading to sub-optimal or non-workable configurations.

Innovation Solution

A method and mechanism for determining resource allocations in computing environments that take into account operational considerations like heat dissipation, avoiding single points of failure, and resource compatibility, using linear programming to select the best allocation based on topology, quality of service requirements, and other criteria.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional random allocation is used, then allocation speed is fast and simplicity is maintained, but operational considerations such as heat dissipation, single points of failure, and resource compatibility are not considered

Engineering Contradiction:

Improveoperational reliabilityVSAvoidallocation mechanism complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system pre-establishes a topology model of the data center infrastructure including rack locations, switch connections, power controller connections, and thermal zones before allocation requests arrive. This preliminary structuring of spatial and operational relationships enables the allocation mechanism to quickly evaluate candidates against operational constraints without complex real-time calculations

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

A linear programming-based allocation mechanism serves as an intermediary between allocation requests and available resources. This intermediary evaluates multiple candidates against operational considerations (heat dissipation, redundancy, compatibility) and selects the optimal allocation, translating complex operational requirements into systematic decision-making without requiring direct complex interactions between all system components

Inventive Principle:

Principle #24Intermediary (Mediator)

2Quantity of substance

If resources are placed physically close to maximize density, then space utilization improves, but heat dissipation problems worsen

Engineering Contradiction:

Improveresource densityVSAvoidheat dissipation

Core Design Contradiction:

Quantity of substanceVSTemperature

Solution Approach 1:

The system divides the data center into thermal zones based on cooling infrastructure and heat dissipation characteristics. Each zone has differentiated thermal properties and cooling capacities. The allocation mechanism assigns resources to specific zones based on their thermal profiles and the thermal requirements of workloads, ensuring that high-density placements occur in zones with adequate cooling capacity rather than uniformly distributing density throughout

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system moves beyond two-dimensional rack space utilization to incorporate vertical and thermal dimensions in allocation decisions. By considering rack elevation positions, airflow patterns, and thermal zones as additional dimensions, the system can place resources in three-dimensional space while respecting thermal constraints, effectively increasing resource density without proportionally increasing heat dissipation problems

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

3Productivity

If resources are shared to maximize utilization, then resource efficiency improves, but single points of failure increase

Engineering Contradiction:

Improveresource utilizationVSAvoidfailure risk

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system proactively identifies and prevents single points of failure by evaluating allocation candidates against redundancy constraints before commitments are made. The linear programming formulation includes constraints that ensure critical infrastructure components (switches, power controllers, cooling units) are not overloaded beyond failure thresholds. This beforehand cushioning against potential failures enables aggressive resource sharing while maintaining reliability bounds

Inventive Principle:

Principle #11Beforehand cushioning (Prior cushioning)

4Productivity

If incompatible resources are allocated to satisfy requests quickly, then allocation speed is maintained, but system compatibility and functionality deteriorate

Engineering Contradiction:

Improveallocation speedVSAvoidresource compatibility

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The system pre-establishes compatibility rules and constraints for resource allocations, including hardware-software compatibility requirements, vendor-specific constraints, and functional compatibility matrices. These compatibility specifications are incorporated into the linear programming formulation before allocation requests are processed, enabling the system to evaluate compatibility as part of the standard optimization process rather than as a separate validation step that would slow down allocation

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS7774467B1Mechanism for making a computing resource allocation

Publication Date: 2010.08.10 ORACLE AMERICAN INC

AI summary

In accordance with one embodiment of the present invention, there are provided methods and mechanisms for determining an allocation of resources, including hardware resources in a computing environment. With these methods and mechanisms, it is possible for computing resource allocations to satisfy one or more operational considerations, such as for example without limitation: “reduce device heat dissipation”, “avoid single point of failure in switched network” and other allocation needs are contemplated.