Transaction Data Filtering for Web Performance Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Web-based applications face challenges in monitoring and identifying performance issues, particularly in determining which portion of the software is responsible for poor performance, leading to customer dissatisfaction and potential sales loss due to unresponsiveness during transactions.
Innovation Solution
The system generates transaction data and applies filtering criteria to create transaction count data and bitmap data, enabling efficient identification of problematic transactions and storage in a database, allowing for querying to determine transaction counts and issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If transaction data is monitored without filtering, then complete transaction visibility is achieved, but data complexity and processing overhead increase significantly
Solution Approach 1:
The patent segments transaction data processing by applying multiple filter criteria (time range, transaction type, status, etc.) to divide the large dataset into manageable subsets. Each filter creates a segmented view of transactions, allowing analysts to focus on specific categories without being overwhelmed by the complete dataset, thus reducing processing complexity while maintaining visibility into relevant transactions.
Solution Approach 2:
The patent extracts specific transaction attributes (error flags, performance metrics, transaction IDs) from the complete transaction data using filtering operations. By extracting only the relevant portions of data that meet specified criteria, the system reduces data complexity while preserving the essential information needed for performance monitoring and diagnostic purposes.
2Measurement precision
If all transaction data is stored and processed, then comprehensive analysis is possible, but storage requirements and processing time increase
Solution Approach 1:
The patent applies filter criteria preliminarily to transaction data before detailed analysis is performed. By pre-filtering transactions based on time ranges, transaction types, and status indicators, the system prepares the data in advance, identifying only those transactions that require further performance analysis. This preliminary action reduces the volume of data subjected to complex performance measurements, thereby decreasing processing time while maintaining analysis accuracy for relevant transactions.
3Productivity
If filtering criteria are applied to reduce data, then processing efficiency improves, but risk of missing problematic transactions increases
Solution Approach 1:
The patent implements multiple independent filter criteria (time range, transaction type, status, error conditions) that can be applied in combination or separately. This multi-functional filtering approach ensures that different categories of problematic transactions can be captured through different filter combinations, maintaining reliable monitoring coverage across various transaction scenarios while preserving processing efficiency through selective filtering.
Solution Approach 2:
The patent incorporates feedback mechanisms where filter results are analyzed to identify problematic transactions, which then inform adjustments to filtering criteria. By continuously monitoring the effectiveness of filters and adjusting them based on observed transaction patterns and performance issues, the system maintains reliable detection of problematic transactions while optimizing processing efficiency through refined filtering strategies.
Data Source
AI summary
The present disclosure relates to a system and method of generating a transaction count using filtering. Transaction data is collected for a set of transactions from an original data set. The collected transaction data is narrowed into a transaction data set by identifying a specific range of the data to be extracted. This transaction data set is filtered against one or more criteria. The transactions included in the transaction data set are then arranged into data packs in a format enabling efficient count generation. The count generation determines how many transactions meet the applied criteria. The data may then be efficiently stored in a database for later retrieval.