In-Memory Virtualization Metrics Aggregation for Cloud Diagnostics

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current techniques for diagnosing issues in virtualization service operations lack efficiency, as they overwhelm storage capacity and are impractical for storing and analyzing execution metrics, leading to undiagnosed performance degradations in cloud computing environments.

Innovation Solution

Implementing a system that generates and aggregates execution metrics for virtualization service operations using in-memory processing, storing them in a dual list structure to facilitate efficient storage, retrieval, and analysis, allowing for faster lookup and better CPU utilization.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If execution metrics for virtualization service operations are stored in traditional storage systems, then diagnostic information is preserved, but storage capacity is overwhelmed and retrieval efficiency deteriorates

Engineering Contradiction:

Improvediagnostic capabilityVSAvoidstorage capacity

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent changes the physical state of metric storage from persistent disk-based storage to volatile in-memory storage. This parameter change enables faster retrieval and processing while reducing the effective quantity of stored data, as memory provides both speed and compression benefits for execution metrics

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent creates compressed representations of execution metrics that capture essential diagnostic information without storing complete raw data. These compressed metric copies are stored in memory, providing diagnostic capability while consuming minimal storage capacity

Inventive Principle:

Principle #26Copying

2Loss of information

If execution metrics are stored and processed using conventional methods, then data is preserved, but processing speed and CPU utilization deteriorate

Engineering Contradiction:

Improvemetric data preservationVSAvoidprocessing speed

Core Design Contradiction:

Loss of informationVSProductivity

Solution Approach 1:

The patent changes the storage medium parameter from disk to memory, which fundamentally improves access speed and processing efficiency. In-memory processing eliminates I/O bottlenecks and enables faster CPU utilization for metric aggregation and analysis operations

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent extracts only the essential diagnostic information from complete execution metrics, storing compressed representations in memory. This extraction maintains the necessary diagnostic capability while enabling faster processing speeds by reducing data volume and complexity

Inventive Principle:

Principle #2Taking out (Extraction)

3Measurement precision

If detailed execution metrics are stored for analysis, then diagnostic precision is improved, but storage requirements and system complexity increase

Engineering Contradiction:

Improvediagnostic precisionVSAvoidstorage and processing complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent changes the storage parameter from persistent disk storage to volatile memory, which provides both faster access and effective data compression. This parameter change maintains diagnostic precision through in-memory data structures while reducing overall system complexity by eliminating complex storage management requirements

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent segments execution metrics into compressed representations that capture only essential diagnostic information. This segmentation maintains measurement precision for critical metrics while reducing storage requirements and simplifying processing complexity by focusing on key performance indicators

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS20230409361A1Generating, aggregating, and querying virtualization service execution metrics using in-memory processing

Publication Date: 2023.12.21 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Techniques for aggregating execution metrics for virtualization service operations are provided. In an example implementation, a command configuring a computer system emulator on a host computer triggers execution of a plurality of virtualization service operations by a virtualization service provider running in a virtualization stack on the host machine. In-memory processing is used to aggregate execution metrics for each type of supported virtualization service operation during a current interval, and at the end of each interval, the execution metrics are pushed to a structure in the memory storing historical aggregated execution metrics. Aggregating and storing execution metrics in-memory enables faster lookup, faster aggregation, and better CPU utilization. Since aggregated metrics are effectively compressed, diagnostic information about a variety of different types of virtualization service operations may be stored and used to diagnose and repair underperforming components.