Reflective Memory Bridge for Database Node Synchronization
Find Innovative SolutionsGenerate 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
Engineering 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
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.
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.
2Productivity
If the number of parallel transactions in the system increases, then system throughput increases, but data contention and metadata contention increase
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.
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.
3Reliability
If k-safety (synchronous replication) is implemented, then data durability is improved, but system complexity and latency increase
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.
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.
4Loss of time
If reflective memory is used for data synchronization, then transaction response time is reduced, but device complexity increases
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.
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.
Data Source
Figure 1
Figure 2A
Figure 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.