Database Storage Tunnels for Mixed Workload Optimization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional relational databases are inefficient in supporting both operational and analytical workloads simultaneously due to differences in normalization approaches, leading to complexity, integration issues, and performance bottlenecks caused by cache pollution and inefficient data retrieval strategies.

Innovation Solution

A data storage and retrieval system utilizing a plurality of data tunnels for each entity type and attribute, with a storage engine that accesses specific data tunnels based on operators and operands, optimizing data retrieval by aligning storage structures with access patterns and reducing cache pollution through efficient data alignment and indexing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional relational databases use normalized schemas for operational workloads, then data update efficiency is improved, but analytical query performance deteriorates due to excessive join operations

Engineering Contradiction:

Improvedata update efficiencyVSAvoidanalytical query performance

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent segments the database system into multiple database instances, each specialized for either operational or analytical workloads. The operational database instance maintains normalized schemas optimized for data updates, while the analytical database instance maintains denormalized schemas optimized for analytical queries. This segmentation allows each instance to be optimized for its specific workload type without compromising the other.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a data replication mechanism as an intermediary between operational and analytical database instances. Data is replicated from the operational instance to the analytical instance, allowing analytical queries to access pre-aggregated, denormalized data without impacting operational data integrity or requiring real-time joins during analytical processing.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Loss of time

If conventional relational databases use denormalized schemas for analytical workloads, then query performance is improved, but data update efficiency deteriorates due to duplication overhead

Engineering Contradiction:

Improvequery performanceVSAvoiddata update efficiency

Core Design Contradiction:

Loss of timeVSProductivity

Solution Approach 1:

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Inventive Principle:

Principle #24Intermediary (Mediator)

3Device complexity

If a single database instance supports both operational and analytical workloads, then system complexity is reduced, but performance bottlenecks occur due to cache pollution and conflicting access patterns

Engineering Contradiction:

Improvesystem complexityVSAvoidperformance

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The patent segments the database system into multiple database instances, each specialized for either operational or analytical workloads. This segmentation isolates cache structures and memory allocations for each workload type, preventing cache pollution where analytical queries would evict operational data from cache and vice versa. Each instance can maintain its own optimization strategies without interfering with the other.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent creates a universal database system architecture that can handle both operational and analytical workloads through multiple specialized instances. The system as a whole provides multi-functionality by supporting different data models, schemas, and access patterns in separate instances, while each instance maintains its own specialized optimizations.

Inventive Principle:

Principle #6Universality (Multi-functionality)

4Productivity

If different database instances are used for operational and analytical workloads, then performance is improved, but integration complexity and data consistency maintenance deteriorate

Engineering Contradiction:

ImproveperformanceVSAvoidintegration complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent introduces a data replication mechanism as an intermediary between operational and analytical database instances. A replication agent monitors data changes in the operational instance and automatically replicates them to the analytical instance, maintaining data consistency without requiring complex manual integration. This intermediary handles the complexity of data synchronization, leaving each instance simple and focused on its specific workload.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent creates a copy of the operational database data in the analytical database instance, allowing the analytical instance to work with replicated data without affecting the operational instance. This copying approach enables independent optimization of each instance while maintaining data consistency through the replication mechanism.

Inventive Principle:

Principle #26Copying

Data Source

PatentUS9965481B2Apparatus, systems and methods for data storage and/or retrieval based on a database model-agnostic, schema-agnostic and workload-agnostic data storage and access models

Publication Date: 2018.05.08 EDGE INTELLIGENCE SOFTWARE INC

AI summary

A database access model and storage structure that efficiently support concurrent OLTP and OLAP activity independently of the data model or schema used, are described. The storage structure and access model presented avoid the need to design schemas for particular workloads or query patterns and avoid the need to design or implement indexing to support specific queries. Indeed, the access model presented is independent of the database model used and can equally support relational, object and hierarchical models amongst others.