Virtual Tag Allocation for Scalable Cloud Resource Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mechanisms for allocating resource utilization, such as cost, in computing clouds rely on inefficient operations like Cartesian products, leading to quadratic complexity and poor scalability, especially when dealing with large datasets.
Innovation Solution
The use of an ASOF JOIN operation to match normalized resource and telemetry values, eliminating the need for full Cartesian products, reduces complexity to O(N log M) and enables scalable, real-time allocation of resource utilization across multiple cloud environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional mechanisms like auto-scaling and load balancing are used to optimize resource utilization, then resource allocation efficiency is improved, but computational complexity increases quadratically due to Cartesian product operations
Solution Approach 1:
The patent transforms the matching problem from timestamp-based to value-based by introducing normalized resource values and telemetry values. This parameter transformation enables the use of efficient range queries instead of exhaustive Cartesian product operations, reducing computational complexity from quadratic to linear or log-linear while maintaining accurate resource-to-entity matching
Solution Approach 2:
The patent introduces an intermediary normalized value scale (0-1 range) that mediates between raw resource metrics and telemetry data. This intermediary representation allows for efficient comparison and matching without requiring direct timestamp alignment, enabling scalable real-time resource allocation across multiple cloud environments
2Measurement precision
If real-time resource allocation is implemented across multiple cloud environments, then resource utilization accuracy is improved, but processing time and computational overhead increase
Solution Approach 1:
The patent performs preliminary normalization of resource values and telemetry values to a common scale before matching operations. This pre-processing step transforms raw data into a standardized format that enables efficient range queries and reduces the computational burden during real-time allocation, allowing accurate matching without excessive processing delays
Solution Approach 2:
The patent replaces the mechanical timestamp-based matching mechanism with a value-based range query system. Instead of sequentially comparing timestamps across multiple clouds, the system uses normalized value ranges to directly identify matching resource-telemetry pairs, significantly reducing processing time while maintaining real-time accuracy
Data Source
AI summary
There is provided a method, comprising: grouping resource records based on respective timestamps and first metadata attributes, each resource record including a resource value indicating utilization of a resource of a computing cloud, adding telemetry records to an indexed dataset separate from the resource records with defined time-based validity periods, each telemetry record including a telemetry value indicating a measurable activity within the computing cloud, each telemetry record is associated with a timestamp and second metadata attributes, normalizing resource values of the resource records and telemetry values of the telemetry records to a defined range representing a common scale, matching the resource records and the telemetry records by matching the normalized resource values and the normalized telemetry values, and generating and assigning virtual tags according to the matches, each virtual tag associating utilization of a specific resource to a specific entity for a specific measurable activity.