VM Migration Biasing to Hyperconverged Nodes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional virtual computing systems face challenges with network congestion and performance degradation due to increased network traffic during virtual machine migration, especially when compute-only nodes lack storage resources, leading to I/O bottlenecks and reduced throughput.
Innovation Solution
A method and system for migrating virtual machines from compute-only nodes to hyperconverged nodes that host their associated virtual disks, optimizing network performance by reducing network load and improving I/O efficiency by biasing VMs to migrate to HC nodes hosting their data storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If VMs are migrated from CO nodes to HC nodes, then I/O performance is improved and network congestion is reduced, but network traffic during migration increases
Solution Approach 1:
The system performs preliminary actions by identifying VMs on CO nodes that are candidates for migration to HC nodes before the actual migration occurs. The management processor proactively determines which VMs should be migrated based on system state, preparing the migration plan in advance to optimize I/O performance while managing network traffic implications.
Solution Approach 2:
The patent introduces an intermediary mechanism - the management processor that mediates between CO nodes and HC nodes. This intermediary identifies and manages the migration process, acting as a controller that balances the trade-off between improving I/O performance by migrating VMs and managing the network traffic generated during migration.
2Productivity
If VMs are co-located on CO nodes, then resource utilization is improved, but network congestion occurs due to increased network traffic during migration
Solution Approach 1:
The management processor implements feedback by monitoring the state of CO nodes and HC nodes, identifying when VMs should be migrated based on system conditions. This feedback mechanism allows the system to respond to changing conditions, migrating VMs from CO nodes to HC nodes to prevent network congestion while maintaining high resource utilization.
Solution Approach 2:
The system applies dynamics by making the VM placement flexible and adaptable. Instead of fixed placement, the management processor dynamically identifies and executes migrations from CO nodes to HC nodes based on real-time system state, allowing the system to optimize resource utilization while preventing network congestion through adaptive reconfiguration.
Data Source
AI summary
A system and method include migrating virtual machines (VMs) between compute only (CO) and hyper converged (HC) nodes. The method includes identifying, by a management processor of a virtual computing system, a plurality of virtual machines hosted on compute only (CO) nodes in the virtual computing system. The management processor then identifies hyper converged (HC) nodes having virtual disks hosting data for the plurality of virtual machines hosted on CO nodes. When a virtual machine (VM) in the plurality of virtual machines is migrating, the management processor biases the VM to migrate to a first HC node hosting a virtual disk assigned to host data for the VM.