In-Memory Transactional Processor for Parallel Business Execution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional transactional processing is limited by sequential execution of business transactions, leading to sub-optimal execution speed and inefficiencies in today's fast-paced business environment, particularly in applications like sales order management where multiple steps require data from previous transactions, causing delays and increased costs.
Innovation Solution
An in-memory database system is linked with applications to enable in-memory-enabled transactional processing, allowing for the instantiation of multiple business transactions with process steps, storing results in an in-memory database for immediate availability, and evaluating alternative process steps at runtime based on contextual data, enabling real-time or near real-time decision-making.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional sequential transactional processing is used, then data consistency is maintained, but processing speed and execution time deteriorate
Solution Approach 1:
The patent pre-loads transactional data into memory before actual processing occurs. The system maintains pre-computed data structures and contextual information in memory, so that when a transaction needs to be processed, the data is already available immediately without requiring disk I/O operations. This preliminary preparation of data in memory significantly accelerates transaction processing speed while maintaining data consistency through proper memory management and transaction protocols.
Solution Approach 2:
The patent replaces traditional disk-based database storage and retrieval mechanisms with in-memory data structures. By substituting mechanical disk I/O operations with electronic memory access, the system achieves dramatically faster data access speeds. The transactional processor directly manipulates data in memory rather than reading from and writing to disk, eliminating the bottleneck of mechanical storage systems while maintaining data integrity through structured memory management.
2Productivity
If multiple business transactions are processed sequentially, then system complexity is reduced, but productivity and throughput deteriorate
Solution Approach 1:
The patent divides the transactional processing system into distinct modular components: a transactional processor, an in-memory database, and multiple parallel transaction execution units. Each transaction can be segmented into independent process steps that can be executed concurrently. This segmentation allows multiple transactions to progress through different stages simultaneously, increasing throughput while keeping each individual processing unit relatively simple and manageable.
Solution Approach 2:
The patent transitions from single-dimensional sequential processing to multi-dimensional parallel processing by adding the time dimension. Multiple transactions are executed simultaneously across different process steps rather than one after another. The in-memory architecture enables this dimensional expansion by providing fast access to data needed for parallel operations, allowing the system to handle multiple transactions at different stages of processing at the same time, thereby dramatically increasing productivity.
3Loss of information
If data is stored in conventional database, then persistence is achieved, but immediate data availability for parallel transactions deteriorates
Solution Approach 1:
The patent introduces an in-memory database as an intermediary layer between the transactional processor and persistent storage. This intermediate memory structure holds recently accessed and frequently used data, allowing parallel transactions to immediately access required information without waiting for disk I/O operations. The in-memory database acts as a buffer that provides fast data availability while eventually synchronizing with persistent storage to ensure data is not lost, thus resolving the contradiction between immediate availability and data persistence.
4Adaptability or versatility
If reactive processing is used, then resource management is simplified, but decision-making speed and responsiveness deteriorate
Solution Approach 1:
The patent implements real-time feedback mechanisms where the transactional processor continuously monitors transaction progress, data changes, and system state in memory. This immediate feedback allows the system to make adaptive decisions about transaction routing, resource allocation, and conflict resolution without waiting for periodic updates or reactive triggers. The feedback loop operates continuously in the memory-based system, enabling fast responsive decision-making that can dynamically adjust to changing conditions while maintaining simplified resource management through centralized control.
Data Source
AI summary
The embodiments provide a system for linking an application with an in-memory database system for enabling an in-memory-enabled transactional application. The system includes an in-memory transactional processor configured to receive contextual data related to an instantiation of a first business transaction of the application and instantiate at least one second business transaction of the application during a course of processing the first business transaction. The in-memory transactional processor is configured to store results of at least one process step of the at least one second business transaction during the course of processing the first business transaction in an in-memory database, evaluate, at a decision point within the first business transaction, a plurality of alternative process steps, during runtime, and determine a subsequent process step from the plurality of alternative process steps based on the contextual data and captured information stored in the in-memory database.