Distributed Database Capacity Planning With Time-Series Throughput Forecasting

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional provisioning methodologies for cloud distributed databases result in significant unused capacity and resource wastage, compounded by their growing popularity and use, leading to inefficient utilization of computing resources.

Innovation Solution

Implementing refined throughput data prediction using time-series analysis and machine learning algorithms to identify and filter out outliers, allowing for dynamic adjustment of database capacity based on predicted usage patterns, thereby minimizing unused resources and optimizing hardware utilization.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional provisioning methodologies are used for cloud distributed databases, then capacity is provisioned in advance, but significant unused capacity and resource wastage occur

Engineering Contradiction:

Improvecapacity availabilityVSAvoidresource wastage

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The system performs preliminary analysis of historical throughput data using time-series analysis and machine learning algorithms to predict future capacity requirements. This allows capacity to be provisioned in advance based on accurate predictions rather than conservative estimates, resolving the contradiction between ensuring capacity availability and avoiding resource wastage

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system continuously monitors actual throughput data and compares it with predicted values, using this feedback to refine predictions and adjust capacity provisioning dynamically. This closed-loop approach ensures capacity matches actual demand while maintaining reliability, eliminating both unused capacity and insufficient capacity scenarios

Inventive Principle:

Principle #23Feedback

2Adaptability or versatility

If conventional provisioning methodologies are used, then capacity is allocated statically, but unused database capacity increases with growing popularity and use

Engineering Contradiction:

Improvecapacity allocation flexibilityVSAvoidunused database capacity

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The system transitions from static capacity allocation to dynamic capacity provisioning by continuously analyzing throughput data patterns and adjusting capacity assignments in real-time. This allows the system to adapt to growing database popularity and usage changes, ensuring capacity matches actual demand and eliminating unused database capacity

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the parameter of capacity allocation from fixed to variable by using machine learning models that continuously update predictions based on historical data patterns. This enables flexible capacity adjustment that grows with database popularity while minimizing unused capacity through precise demand forecasting

Inventive Principle:

Principle #35Parameter changes

3Quantity of substance

If outlier data is included in throughput analysis, then more data points are available for prediction, but prediction accuracy decreases

Engineering Contradiction:

Improvedata volumeVSAvoidthroughput prediction accuracy

Core Design Contradiction:

Quantity of substanceVSMeasurement precision

Solution Approach 1:

The system extracts and removes outlier data points from the throughput data before performing time-series analysis and prediction. By filtering out anomalies and extreme values, the system maintains high prediction accuracy while still utilizing the majority of valid historical data points for pattern recognition and forecasting

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system introduces an intermediary outlier detection and filtering layer between data collection and prediction analysis. This intermediary component uses statistical methods and machine learning to identify and remove outliers, allowing the system to process large volumes of data while maintaining high prediction accuracy through clean, validated input data

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS12596587B2Cloud distributed database capacity planning and adjustment using time-series data analysis

Publication Date: 2026.04.07 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Systems and methods are provided for implementing cloud distributed database capacity planning and/or adjustment, using time-series data analysis. In various embodiments, a computing system may be used to analyze collected throughput data associated with consumption of provisioned throughput resources of a distributed cloud database over one or more past periods by an entity. Based on the analysis, a set of predicted throughput data may be determined or generated for the entity over a future upcoming period. In some cases, based on a determination that adding physical partitions would be required, the computing system may adjust the set of predicted throughput data to reduce or minimize a number of physical partitions to be added. The provisioned capacity of the distributed cloud database may then be dynamically adjusted based at least in part on one of the unadjusted or adjusted set of predicted throughput data.