Dual System Storage Mode for In-Memory Database Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current enterprise resource planning (ERP) systems face issues with data replication and synchronization between source and in-memory systems, leading to multiple sources of truth, time lags, data inconsistencies, and limitations in writing transactions directly to in-memory systems, which hinder migration to in-memory systems.
Innovation Solution
Implementing a dual system storage mode that uses a lookup table to route queries and updates between the original database and in-memory database, allowing direct writing and real-time updates in the in-memory system without replicating data back to the ERP system, thus maintaining a single source of truth and reducing delays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If data is replicated from ERP system to in-memory system, then analytics processing speed is improved, but data consistency between systems deteriorates due to multiple sources of truth
Solution Approach 1:
The patent introduces a dual system storage mode that acts as an intermediary layer between the ERP system and in-memory system. This mode includes a storage mode indicator that tracks which system is the current source of truth, and a query routing mechanism that directs read/write operations to the appropriate system, thereby maintaining data consistency while enabling fast analytics.
Solution Approach 2:
The patent implements dynamic switching between storage modes (ERP system storage mode and dual system storage mode). The system can transition from static replication to dynamic operation where the source of truth can change based on operational needs, allowing flexible control over data consistency and analytics performance.
2Reliability
If replication tools are used to propagate changes, then data synchronization is maintained, but time lags and delays between systems increase
Solution Approach 1:
The patent extracts the data synchronization function from traditional replication tools and integrates it directly into the database system through the dual system storage mode. This eliminates the intermediate replication layer that causes time lags, allowing real-time synchronization between ERP system and in-memory system.
3Productivity
If direct writing to in-memory system is enabled, then analytics performance is improved, but data inconsistency and locking complications increase
Solution Approach 1:
The patent implements a feedback mechanism through the storage mode indicator that tracks the current state of data storage. When writing operations occur, the system updates the storage mode indicator to reflect the new source of truth, providing continuous feedback that prevents conflicting write operations and reduces locking complexity.
4Adaptability or versatility
If migration to in-memory system is pursued, then analytics capability is improved, but user reluctance increases due to data consistency concerns
Solution Approach 1:
The patent creates a universal system that can operate in multiple modes (ERP system storage mode and dual system storage mode) and support multiple data sources. This multi-functionality allows users to gradually migrate to the in-memory system while maintaining the ability to fall back to the ERP system, reducing user reluctance through flexible adoption.
Data Source
AI summary
A system and method for running an application on multiple databases is described. A selection of data stored in a database to be used in conjunction with an in-memory database is identified. A selection of a storage mode corresponding to a dual system storage mode or a source system storage mode is received. The selection of data is initialized in the in-memory database. An entry corresponding to the selection of data is generated in a lookup table when the dual system storage mode is selected.