Dynamic Virtual Machine Provisioning via Network Bandwidth Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In cloud computing environments, virtual machines often experience network bandwidth over-subscription and increased latency due to uneven distribution of network resources across compute nodes, leading to reduced response times and inefficient resource utilization.
Innovation Solution
A method that manages virtual machines by migrating them from compute nodes with over-utilized Ethernet links to those with under-utilized links, using data from a network switch's management information database to determine bandwidth usage and provision additional virtual machines to nodes with the greatest unutilized bandwidth, thereby balancing network bandwidth and ensuring each virtual machine receives the required resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If virtual machines are assigned to compute nodes based on storage requirements only, then storage allocation efficiency is improved, but network bandwidth utilization becomes unbalanced leading to over-subscription and increased latency
Solution Approach 1:
The patent implements dynamic provisioning that continuously monitors network bandwidth utilization and dynamically assigns virtual machines to compute nodes based on real-time network conditions. The system periodically evaluates network utilization metrics and redistributes virtual machines from over-subscribed nodes to under-subscribed nodes, transforming the static storage-based allocation into a dynamic resource-balancing system that adapts to changing network conditions.
2Productivity
If virtual machines are concentrated on fewer compute nodes, then resource consolidation is improved, but network latency increases due to bandwidth over-subscription
Solution Approach 1:
The patent changes the allocation parameters from static storage-based metrics to dynamic network-utilization-based metrics. The system monitors network bandwidth usage as a key parameter and uses this information to determine virtual machine placement, thereby adjusting the allocation strategy based on real-time network conditions rather than fixed storage requirements alone.
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously monitors network bandwidth utilization and uses this feedback to make informed decisions about virtual machine provisioning and migration. The monitoring data feeds back into the provisioning logic, enabling the system to respond to network conditions and adjust resource allocation to maintain optimal performance.
3Device complexity
If network bandwidth is not monitored and balanced, then system complexity is reduced, but resource utilization efficiency deteriorates leading to performance degradation
Solution Approach 1:
The patent enables the provisioning system to automatically monitor its own network conditions and self-correct imbalances by dynamically provisioning and migrating virtual machines. The system serves itself by autonomously detecting network utilization issues and implementing corrective actions without external intervention, thereby maintaining high resource utilization efficiency while managing complexity through automation.
Data Source
AI summary
Virtual machines are managed on the basis of data obtained from a management information database of a network switch having a plurality of Ethernet links coupled to compute nodes running a plurality of virtual machines. A management entity, such as a provisioning manager, determines the amount of network bandwidth being utilized through each of the Ethernet links. Accordingly, an additional virtual machine may be provisioned to the compute node that is coupled to an Ethernet link identified as having the greatest amount of unutilized network bandwidth.