Dynamic VM Placement and Buffer Tuning for Seasonal Cloud Load

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Data centers face inefficiencies due to poor network performance caused by inefficient allocation and configuration of resources, leading to lost data, customer dissatisfaction, and cost inefficiencies, particularly due to seasonal variations in network throughput that current static implementations fail to address.

Innovation Solution

The implementation of machine learning-based techniques to forecast seasonal network bandwidth oversubscription and dynamically adjust network device resources, such as buffer sizes, and migrate virtual machines to optimize resource utilization across clustered systems, ensuring efficient network performance and reducing packet loss.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If static network device configuration is used, then device complexity is reduced, but network performance deteriorates during seasonal load variations

Engineering Contradiction:

Improvenetwork device configurationVSAvoidnetwork performance

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The network device configuration transitions from static to dynamic through automated seasonal adjustment. The system monitors network load patterns, identifies seasonal variations, and automatically adjusts buffer sizes and resource allocation accordingly. This dynamic adaptation resolves the contradiction by allowing the configuration to change based on actual network conditions while maintaining automated control.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The invention changes key network device parameters (buffer sizes, resource allocation) based on detected seasonal patterns. By analyzing historical network performance data and identifying recurring seasonal load variations, the system automatically adjusts configuration parameters to match predicted demand, thereby maintaining performance without requiring complex manual reconfiguration.

Inventive Principle:

Principle #35Parameter changes

2Reliability

If network device buffers are maximized to reduce packet loss, then reliability improves, but system resource efficiency deteriorates during low activity periods

Engineering Contradiction:

Improvepacket loss reductionVSAvoidsystem resource efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system implements periodic adjustment of buffer sizes based on seasonal network load patterns. Instead of maintaining maximum buffers continuously, the system detects seasonal variations and adjusts buffer allocation periodically to match predicted demand. This resolves the contradiction by providing sufficient buffers during high-demand seasonal periods while reducing buffer allocation during low-activity periods, thereby maintaining reliability when needed and improving resource efficiency when demand is low.

Inventive Principle:

Principle #19Periodic action

3Reliability

If virtual machines are migrated to optimize network throughput, then network performance improves, but device complexity increases

Engineering Contradiction:

Improvenetwork throughputVSAvoidVM management complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system implements self-service automated VM migration based on seasonal network load detection. The hypervisor monitors network performance metrics, identifies seasonal patterns, and automatically migrates VMs to optimize throughput without manual intervention. This automated self-service approach resolves the contradiction by handling the complexity of VM management internally while delivering improved network performance, shielding users from the underlying complexity.

Inventive Principle:

Principle #25Self-service

4Adaptability or versatility

If manual intervention is used to adjust network device settings, then adaptability improves, but loss of time increases due to reset requirements

Engineering Contradiction:

Improveconfiguration adaptabilityVSAvoiddevice reset time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system performs preliminary automated detection and adjustment of network device settings based on seasonal patterns before performance degradation occurs. By monitoring historical data and predicting seasonal load variations, the system proactively adjusts configurations in advance, eliminating the need for time-consuming manual resets and device downtime while maintaining adaptability to changing conditions.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11997159B2Proactive placement of virtualized computing resources and tuning of network resources based on seasonal variations

Publication Date: 2024.05.28 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Techniques are disclosed for configuring network device resources and determining locality of virtual machines in a virtual network of a cloud computing environment. Usage data for computing resources in the computing network is collected. Based on the collected usage data, a time-based usage profile for the computing resources in the computing network is determined. A periodic component in the time-based usage profile is determined. Based on the periodic component, distribution of VMs is determined or the buffer size of network devices is configured.