Database Key Value Distribution via Logical Partitioning

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Latch contentions occur in database management systems due to skewed distributions of key values, leading to exclusive access issues and increased waiting times for multiple processes attempting to access the same nodes.

Innovation Solution

Distributing key values by generating a logical partition associated with a partition value, which is attached to the key values to create an altered key value, thereby reducing latch contentions and optimizing access to data structures like B-plus trees.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If key values are distributed unevenly (skewed distribution), then data structure organization reflects the data naturally, but latch contentions increase and access performance deteriorates

Engineering Contradiction:

Improvedata structure consistencyVSAvoidaccess speed

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent segments the key value space by introducing a partition value that divides keys into different ranges. This segmentation distributes the key values across multiple partitions, preventing concentration at single nodes and thereby reducing latch contentions while maintaining data structure integrity.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent modifies the key value parameter by attaching a partition value to it, creating an altered key value. This parameter transformation changes the distribution characteristics of keys, steering them away from skewed concentrations toward a more balanced distribution across different partitions.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If multiple processes access the same node simultaneously, then concurrency is utilized, but latch contentions occur and waiting time increases

Engineering Contradiction:

Improveconcurrency utilizationVSAvoidwaiting time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

By segmenting the data structure into multiple partitions based on altered key values, the patent enables multiple processes to access different partitions simultaneously without conflicting for the same latch. This segmentation reduces waiting time while maintaining high concurrency utilization.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The partition value acts as an intermediary that redirects access requests to different partitions. Instead of multiple processes competing for the same node, the intermediary distribution mechanism routes them to different nodes, eliminating latch contentions and reducing waiting time.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Productivity

If logical partitions are introduced to distribute key values, then latch contentions are reduced, but data structure complexity increases

Engineering Contradiction:

Improveaccess performanceVSAvoiddata structure complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent introduces complexity in a controlled manner by modifying only the key value parameter through attachment of partition values. This parameter-based approach adds logical partitioning functionality without fundamentally restructuring the entire data structure, thus managing complexity while improving access performance.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS8738657B2Distribution of key values

Publication Date: 2014.05.27 HEWLETT PACKARD ENTERPRISE DEV LP

AI summary

A computer apparatus and related method to reduce database congestion is provided. In one aspect, the computer apparatus and method may generate a new partition within a hierarchical data structure of interlinked nodes, if a distribution of key values stored in the hierarchical data structure is skewed in favor of a range of key values.