Time-series Data Accumulation with Interval Function Determination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data accumulation systems face challenges in retaining old data for analysis, as they often delete data below a significance threshold, preventing consideration of old data's influence in analysis processes.
Innovation Solution
A time-series data accumulation device comprising a storage unit, determination unit, calculation units, and a write unit, which determines an interval function based on data capacity and elapsed time to calculate and store statistical values, allowing for the inclusion of old data in analysis without deletion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is accumulated in a database of fixed capacity by defining a significance threshold increasing with elapsed time, then the database capacity is efficiently utilized, but old data below the significance threshold is deleted and cannot be used for data analysis
Solution Approach 1:
The database is divided into multiple storage regions (first storage region for recent data, second storage region for old data). This segmentation allows different data to be stored with different retention policies, enabling both efficient capacity utilization and preservation of old data for analysis.
Solution Approach 2:
Different storage regions are assigned different quality characteristics - the first storage region maintains high-quality recent data with fine time intervals, while the second storage region preserves older data with coarser time intervals. This local quality differentiation resolves the contradiction by allowing efficient space usage while retaining analytically valuable old data.
2Quantity of substance
If old data is moved to a tape device for lower speed processing, then database capacity is freed up, but data analysis requiring old data requires restoration from tape device
Solution Approach 1:
Old data is pre-processed and stored in the second storage region with aggregated statistical values before being needed for analysis. This preliminary action eliminates the need for time-consuming restoration from tape devices, as the data is already prepared and accessible in the database.
Solution Approach 2:
Instead of moving original detailed data to tape, aggregated statistical copies of old data are stored in the second storage region. These copies preserve essential analytical information while occupying minimal space, eliminating restoration needs while maintaining database capacity.
3Adaptability or versatility
If data is imported from another database for analysis, then necessary data can be obtained, but data must be imported again when analyzing different data items
Solution Approach 1:
The database is designed to store multiple types of data items (recent detailed data and old aggregated data) in a unified structure with different storage regions. This universal design allows any data analysis request to be fulfilled directly from the database without repeated imports, serving multiple analysis needs simultaneously.
Data Source
AI summary
A storage unit of a time-series data accumulation device of an embodiment stores statistical value data. A determination unit determines an interval function on the basis of data capacity, and a total amount of time elapsed from the beginning of measurement of a time-series data. A first calculation unit calculates an interval indicating an amount of time required to measure all the time-series data items constituting desired statistical value data on the basis of the determined interval function. A second calculation unit acquires all the time-series data items on the basis of the calculated interval, and thereafter calculates the statistical value on the basis of all the acquired time-series data items. A write unit writes the calculated statistical value to the storage unit as the statistical value data.