Dynamic Weight Assignment for Client Placement in Computer Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Efficient placement of resource-consuming clients like virtual machines in computer network systems is challenging due to the complexities of hosting and storage device attributes, such as CPU, memory, network bandwidth, free space, latency, and IOPS, which require balancing conflicting goals and varying resource utilization metrics.
Innovation Solution
A system and method using continuously variable weights for resource utilization metrics to compute selection scores for candidate devices, enabling the selection of an optimal target device for client placement, balancing resource utilization across clusters and datastores to maximize efficient resource allocation and minimize load imbalance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple resource utilization metrics (CPU, memory, network bandwidth, free space, latency, IOPS) are considered for client placement, then placement accuracy and resource optimization are improved, but system complexity and difficulty of making placement decisions increase
Solution Approach 1:
The patent transforms multiple resource utilization metrics into a single composite placement score by applying dynamic weight adjustments. Different weights are assigned to various metrics (CPU, memory, network bandwidth, free space, latency, IOPS) based on current system conditions and client requirements, allowing the system to handle multiple parameters through a unified scoring mechanism that simplifies the placement decision process while maintaining high accuracy.
2Adaptability or versatility
If static weights are used for resource metrics, then placement decisions are simpler to implement, but the system cannot adapt to varying resource utilization patterns and conflicting goals
Solution Approach 1:
The patent implements dynamic weight adjustments where the weights assigned to different resource metrics are not fixed but vary based on current system state and placement goals. The system can adjust weights in real-time to prioritize certain resources over others depending on the specific client requirements and available resource conditions, enabling adaptability while maintaining operational simplicity through automated weight calculation.
Solution Approach 2:
The system changes the parameter of weights dynamically based on resource utilization patterns and placement objectives. By adjusting weight values according to current system conditions, the patent enables the placement algorithm to adapt to varying resource patterns and conflicting goals without requiring manual intervention or complex rule-based decision-making.
3Productivity
If resource utilization metrics are normalized and weighted to compute selection scores, then optimal client placement is achieved, but computational complexity increases
Solution Approach 1:
The patent transforms multiple resource metrics into a single composite score through normalization and weighted aggregation. By converting diverse metrics (CPU percentage, memory GB, network Mbps, free space TB, latency ms, IOPS) into a unified scoring system with dynamic weights, the system achieves efficient placement decisions while managing computational complexity through mathematical transformation rather than complex algorithmic processing.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system and method for placing a client in a computer network system uses continuously variable weights to resource utilization metrics for each candidate device, e.g., a host computer. The weighted resource utilization metrics are used to compute selection scores for various candidate devices to select a target candidate device for placement of the client.