Database Tool Segments Read and Write Workloads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database systems face a performance trade-off where they cannot be both read-efficient and write-efficient simultaneously, leading to resource-intensive multiple table joins during read operations and substantial resource consumption during write operations.
Innovation Solution
A database tool that maintains the same data in both a write-efficient and a read-efficient database by assigning processors to different transaction types and segments, using a machine learning algorithm to determine which fields to read and send based on user device channels, thereby optimizing processor resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a normalized database model is used to write data to minimal locations, then write efficiency is improved, but read operations require multiple table joins which consume high resources
Solution Approach 1:
The patent divides the database system into two separate databases: a write-efficient database using normalized model and a read-efficient database using denormalized model with pre-computed joins. This segmentation allows each database to optimize for its specific operation type without compromise
Solution Approach 2:
The patent introduces a mediator component that automatically manages data synchronization between the write-efficient and read-efficient databases. This intermediary handles the complex task of translating writes from one database to the other, allowing both databases to maintain their optimal structures
2Productivity
If data is stored in redundant locations to avoid table joins, then read efficiency is improved, but write operations consume substantial resources
Solution Approach 1:
The patent segments the storage system into two distinct databases with different structural characteristics. The read-efficient database uses denormalized storage with redundant data to enable fast reads without joins, while the write-efficient database uses normalized storage to simplify write operations
Solution Approach 2:
The patent creates a copy of the data model in two different forms: one optimized for writing (normalized) and one optimized for reading (denormalized). The mediator ensures that data written to one database is appropriately copied and transformed in the other database
3Device complexity
If a single database is used for both read and write operations, then device complexity is reduced, but processor bottlenecks occur during simultaneous operations
Solution Approach 1:
The patent segments the database workload into two separate systems, each optimized for specific operation types. This segmentation eliminates processor bottlenecks by allowing read and write operations to proceed independently without competing for the same processing resources
Solution Approach 2:
The mediator acts as an intermediary that coordinates between the two databases, managing data consistency and synchronization without creating processor bottlenecks. It allows both databases to operate independently while maintaining data integrity
Data Source
AI summary
A system includes a database, a memory, and a processor. The database stores a set of transactions, each of which includes a set of fields. The processor receives a query from a user through a channel. The query is directed to a transaction stored in the database. The transaction includes a set of fields. The processor applies a machine learning policy adapted to determine a view, i.e., a subset of the set of fields that the user is likely to access using the channel. The processor further accesses the database and retrieves the view. The processor then sends the view to the channel.