Reflective Memory Bridge for Database Node Synchronization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Database systems face challenges with data contention and metadata contention in online transaction processing (OLTP) systems, particularly as the number of parallel transactions increases, leading to higher transaction response times.

Innovation Solution

Employing reflective memory in a multi-node computing and database environment using a reflective memory bridge that maps incoming and outgoing memory regions across nodes, enabling coherent data replication without I/O operations and reducing latency, thus maintaining k-safety without single points of failure.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If traditional disk-based single server systems are used, then data storage capacity is limited, but I/O rates are high and memory requirements are small

Engineering Contradiction:
Improvedata storage capacityVSAvoidI/O rate
Core Design Contradiction:
Quantity of substanceVSProductivity

Solution Approach 1:

The system segments data storage across multiple independent nodes in a cluster, with each node holding a portion of the database in memory. This horizontal partitioning allows unlimited scaling of storage capacity while maintaining high I/O performance through parallel access to multiple nodes.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from a single-server vertical architecture to a multi-node horizontal cluster architecture. This dimensional change allows the system to scale storage capacity linearly by adding nodes while maintaining consistent I/O performance through parallel processing.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If the number of parallel transactions in the system increases, then system throughput increases, but data contention and metadata contention increase

Engineering Contradiction:
Improvesystem throughputVSAvoiddata contention and metadata contention
Core Design Contradiction:
ProductivityVSObject-affected harmful factors

Solution Approach 1:

The database is horizontally partitioned across multiple independent nodes, with each node managing its own local data and metadata. This segmentation isolates contention to local scopes rather than system-wide, allowing parallel transactions to proceed concurrently on different nodes without global contention.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces reflective memory as an intermediary mechanism that enables efficient coordination between nodes. Reflective memory provides a shared view of data across nodes while maintaining local ownership, reducing metadata contention through a consistent read-modify-write protocol.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If k-safety (synchronous replication) is implemented, then data durability is improved, but system complexity and latency increase

Engineering Contradiction:
Improvedata durabilityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

Each node autonomously maintains its own copy of the database and independently handles transactions. The reflective memory mechanism automatically propagates changes to other nodes without requiring complex coordination protocols, simplifying the implementation of k-safety.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system implements synchronous replication by copying data to multiple reflective memory regions across different nodes. Each node maintains an independent copy of the database, and changes are propagated to all copies atomically through the reflective memory mechanism, ensuring k-safety without complex coordination.

Inventive Principle:
Principle #26Copying

4Loss of time

If reflective memory is used for data synchronization, then transaction response time is reduced, but device complexity increases

Engineering Contradiction:
Improvetransaction response timeVSAvoiddevice complexity
Core Design Contradiction:
Loss of timeVSDevice complexity

Solution Approach 1:

The patent replaces traditional mechanical I/O operations with reflective memory operations. Instead of using disk I/O or network protocols for data synchronization, the system uses reflective memory to directly copy data between nodes, eliminating I/O latency and reducing transaction response time.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system changes the fundamental parameter of data storage from disk-based to memory-based, and from sequential access to parallel reflective access. This parameter change enables simultaneous read-modify-write operations across multiple nodes without the latency penalties of traditional I/O systems.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2845110B1Reflective memory bridge for external computing nodes
Publication Date: 2018.10.24 HEWLETT PACKARD ENTERPRISE DEV LP
  • EP2845110B1 patent drawingFigure 1
  • EP2845110B1 patent drawingFigure 2A
  • EP2845110B1 patent drawingFigure 2B

AI summary

In at least some examples, a computing node includes a processor and a local memory coupled to the processor. The computing node also includes a reflective memory bridge coupled to the processor. The reflective memory bridge maps to an incoming region of the local memory assigned to at least one external computing node and maps to an outgoing region of the local memory assigned to at least one external computing node.