Adaptive Resource Offer Rates Using Utilization Monitoring Windows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately determine and manage the offer rate of computing device resources, leading to inefficient data flow and resource utilization, particularly when historic data does not reflect current utilization rates, resulting in overwhelming entities with excessive offers or underutilization.
Innovation Solution
A resource matching platform that periodically determines a mean utilization rate and variability, adjusting the offer rate based on current utilization patterns by selecting optimal monitoring windows, using a combination of mean and adjusted rates to tailor offers to individual entities, and applying throttle rates to manage resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the offer rate is increased to ensure resource availability, then resource utilization improves, but entities are overwhelmed with excessive offers leading to increased infrastructure costs
Solution Approach 1:
The offer rate is made dynamic by adjusting it based on the monitored utilization rate. The system continuously monitors the utilization rate and adapts the offer rate accordingly, transitioning from static to dynamic control to optimize resource allocation while preventing overwhelming entities with excessive offers.
Solution Approach 2:
The system implements feedback control by monitoring the utilization rate and using it to adjust the offer rate. The monitored utilization rate feeds back into the offer rate determination process, creating a closed-loop control system that automatically balances resource availability with entity workload.
2Ease of operation
If historic data is used to determine offer rates, then resource allocation is simplified, but accuracy deteriorates when historic data does not reflect current utilization rates
Solution Approach 1:
The system transitions from static historic data to dynamic real-time monitoring. By continuously monitoring the utilization rate and comparing it against historic data, the system adapts to changing conditions while maintaining the simplicity of rule-based allocation. The dynamic adjustment ensures accuracy even when historic data becomes outdated.
Solution Approach 2:
The system performs preliminary monitoring and evaluation of utilization patterns before finalizing offer rates. By proactively tracking utilization rates and identifying trends, the system prepares adjusted offer rates in advance, ensuring accurate resource allocation when实际需求 arises without relying solely on potentially outdated historic data.
3Measurement precision
If the monitoring window is extended to capture more data, then measurement accuracy improves, but the system includes outdated data that does not reflect current utilization rates
Solution Approach 1:
The system performs preliminary analysis of multiple candidate monitoring windows with different time spans. By evaluating windows of various durations and comparing their results, the system identifies the optimal window length that captures sufficient data for accurate measurement while excluding outdated information. This preliminary evaluation enables the system to adaptively select the most relevant time period.
Solution Approach 2:
The system changes the monitoring window parameter dynamically by evaluating multiple candidate window lengths and selecting the optimal one. Instead of using a fixed window size, the system adjusts the window parameter based on the characteristics of the utilization data, ensuring that the selected window provides accurate current utilization rates without including excessive outdated data.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Resource matching platforms and methods are disclosed. In an example, processing circuitry comprises a monitoring module configured to periodically determine a mean utilisation rate, wherein the mean utilisation rate comprises a rate at which computing device resource is utilised by an entity; and an offer module configured to offer computing device resource to the entity, wherein the offer module is further configured to adapt a rate at which computing device resource is offered to the entity based on a current determined mean utilisation rate for the entity, wherein the mean utilisation rate is by determining (i) a mean utilisation rate and (ii) a measure of variability of the utilisation rate for each of a plurality of candidate monitoring windows, selecting a candidate monitoring window as the monitoring window; and determining the mean utilisation rate for that window as the determined mean utilisation rate for the entity.