Host Controller VM Migration for CPU Load Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Load imbalance in host clusters of virtual machines leads to performance degradation, as some hosts become overloaded while others remain underutilized, necessitating effective load management strategies.
Innovation Solution
A host controller implements a method for live migration of virtual machines from overloaded hosts to underutilized ones, selecting virtual machines based on CPU usage and balancing CPU load across hosts, while considering factors like memory availability and power saving.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If virtual machines are migrated from overloaded hosts to underutilized hosts, then load balance is improved, but system complexity increases
Solution Approach 1:
A host controller is introduced as an intermediary component that centralizes the decision-making logic for VM migration. The host controller receives load information from multiple hosts, determines migration decisions, and coordinates the migration process, thereby managing system complexity in a centralized manner while improving load balance across the cluster.
Solution Approach 2:
The system implements continuous feedback mechanisms where hosts report their load information (CPU usage, memory availability) to the host controller. The host controller uses this feedback to dynamically adjust migration decisions, selecting VMs for migration based on current load conditions and monitoring the effects of migrations to maintain optimal load balance.
2Reliability
If live migration is performed to maintain continuous operation, then service availability is improved, but resource overhead increases
Solution Approach 1:
The system performs partial migration actions by selecting only specific virtual machines for migration based on load thresholds and criteria, rather than migrating all VMs. This partial action approach maintains service availability for critical VMs while reducing the overall resource overhead associated with migration operations.
Solution Approach 2:
The host controller changes migration parameters dynamically based on system conditions, including selecting VMs with the least impact on service availability, adjusting migration timing, and modifying resource allocation parameters during migration to minimize overhead while maintaining continuous service operation.
Data Source
AI summary
A host controller migrates live virtual machines from one host to another host to balance the load in the host cluster. In one embodiment, the host controller identifies an overloaded host from the host cluster. The host controller selects a virtual machine from the existing virtual machines that run on the overloaded host based on a percentage value associated with the virtual machine. The percentage value represents CPU usage incurred by the virtual machine on the host. In one embodiment, the selected virtual machine is one that incurs the least amount of load on the host. The host controller then selects a target host from the host cluster as a destination for migrating the selected virtual machine.