Database Migration Framework Using Machine Learning Models

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Migrating databases to cloud computing systems is challenging due to the difficulty in determining the appropriate characteristics, such as size and configuration, that balance performance and cost, especially for large enterprises with multiple databases, as incorrect estimations can lead to wasted resources or performance issues.

Innovation Solution

A framework utilizing two machine learning models, a neural network-based autoencoder and classifier, generates compressed representations of database characteristics under workloads and provides recommendations for cloud database size, configuration, and service level, facilitating efficient migration planning by creating a similarity space and selecting relevant subsets of performance data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If cloud database size is estimated too high, then performance requirements are met, but cost increases due to wasted resources

Engineering Contradiction:
ImproveperformanceVSAvoidcost
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The system performs preliminary analysis of database workloads and characteristics before migration to cloud, using machine learning models to predict optimal cloud database size and configuration. This advance planning prevents both over-provisioning (wasting resources) and under-provisioning (insufficient performance) by determining the precise requirements beforehand.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system analyzes multiple parameters including database size, configuration options, and workload characteristics to determine the optimal cloud database setup. By changing and optimizing these parameters based on predictive analytics, the system achieves cost-effective resource allocation that meets performance requirements without waste.

Inventive Principle:
Principle #35Parameter changes

2Quantity of substance

If cloud database size is estimated too low, then cost is reduced, but performance issues occur

Engineering Contradiction:
ImprovecostVSAvoidperformance
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The system performs preliminary analysis of database workloads and characteristics before migration to cloud, using machine learning models to predict optimal cloud database size and configuration. This advance planning prevents both over-provisioning (wasting resources) and under-provisioning (insufficient performance) by determining the precise requirements beforehand.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses feedback from analyzing database performance metrics, workload patterns, and migration outcomes to continuously improve its predictions. By incorporating feedback loops that learn from actual performance data, the system becomes more accurate in estimating optimal cloud database size, ensuring sufficient performance while minimizing cost.

Inventive Principle:
Principle #23Feedback

3Adaptability or versatility

If multiple databases are migrated to cloud, then enterprise capabilities are enhanced, but complexity of determining proper characteristics increases

Engineering Contradiction:
Improveenterprise capabilitiesVSAvoidcomplexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system creates a universal framework that handles multiple database migration scenarios through a single integrated machine learning approach. The same predictive models and analysis methods are applied across different databases, configurations, and cloud targets, providing consistent guidance for enterprise-scale migrations without requiring separate complex processes for each database.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system segments the complex migration planning process into manageable components: workload analysis, characteristic prediction, configuration optimization, and cost estimation. By breaking down the overall migration task into these discrete analytical steps, the system reduces complexity while maintaining comprehensive coverage of all necessary considerations for multi-database migrations.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3814900B1Framework for providing recommendations for migration of a database to a cloud computing system
Publication Date: 2025.01.15 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP3814900B1 patent drawingFigure 1
  • EP3814900B1 patent drawingFigure 2
  • EP3814900B1 patent drawingFigure 3

AI summary

To obtain one or more recommendations for the migration of a database to a cloud computing system, information about performance of the database operating under a workload may be obtained. A first machine learning model (e.g., a neural network-based autoencoder) may be used to generate a compressed representation of characteristics of the database operating under the workload. The compressed representation may then be provided as input to a second machine learning model (e.g., a neural network-based classifier), which outputs a recommendation regarding a characteristic (e.g., size, configuration, level of service) of the cloud database to which the database should be migrated. This type of recommendation may be made prior to migration, thereby making it easier to properly estimate the cost of running the cloud database and plan the migration accordingly.