HCI Management Node Co-locating VMs and Volumes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In hyper-converged infrastructure (HCI) environments, the management node's inability to effectively associate virtual machines (VMs) with storage volumes leads to degraded I/O performance and inefficient resource allocation due to bandwidth bottlenecks when VMs and volumes are not optimally arranged across nodes.
Innovation Solution
A computer system with a management device that allocates resources among nodes based on resource ratio differences to ensure VMs and volumes are placed on the same node, optimizing resource utilization and maintaining I/O performance by determining the optimal node for both compute and storage resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the management node arranges VMs and volumes independently without association, then the arrangement process is simple, but I/O performance degrades due to bandwidth bottlenecks
Solution Approach 1:
The patent combines the VM arrangement process with the volume assignment process into a single integrated operation. The management node simultaneously determines both the compute node for the VM and the storage node for the volume, ensuring they are properly associated while maintaining arrangement simplicity.
Solution Approach 2:
The patent introduces an intermediary mechanism that coordinates between compute resource allocation and storage resource allocation. This intermediary process ensures that VMs and volumes are arranged in an associated manner, preventing bandwidth bottlenecks while maintaining operational simplicity.
2Adaptability or versatility
If VMs and volumes are arranged independently, then resource allocation is flexible, but resource utilization becomes inefficient due to bandwidth bottlenecks
Solution Approach 1:
The patent merges the flexibility of independent arrangement with the efficiency of associated arrangement by implementing a coordinated allocation process. The management node maintains flexibility in selecting compute and storage resources while ensuring optimal pairing between VMs and volumes to maximize resource utilization efficiency.
3Device complexity
If VMs and volumes are arranged without association, then the arrangement method is simple, but bandwidth bottlenecks occur leading to performance degradation
Solution Approach 1:
The patent combines VM arrangement and volume assignment into a unified process that maintains simplicity while ensuring proper association. The management node executes a single coordinated arrangement operation rather than separate independent operations, preventing bandwidth bottlenecks without significantly increasing system complexity.
Data Source
AI summary
The present disclosure provides a computer system in which a virtual computer and a volume can be arranged in a cluster without degrading I/O performance of the virtual computer. A management node is configured to arrange a virtual computer and a volume in which the virtual computer inputs and outputs data in any one of a plurality of nodes, so as to allocate a predetermined resource among resources of a processor, a memory, and a storage drive to the virtual computer and the volume, thereby operating the virtual computer and the volume in the node. When the virtual computer and the volume can be arranged in the same node among the plurality of nodes, the management node determines the same node from the plurality of nodes based on a difference of ratios of resource amounts allocated to the virtual computer and the volume between the plurality of resources.