Performance Impact Evaluation for Cloud Data Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Migrating communication systems to a cloud environment faces challenges in achieving operational efficiency and resource utilization due to individuality of processing servers, and the additional access to an external DB extends call processing Turn Around Time (TAT), making it difficult to accurately evaluate performance impacts during data separation.
Innovation Solution
A performance impact evaluation apparatus that acquires logs of access to an external DB and evaluates performance impact using specific evaluation expressions considering software structure and data characteristics, distinguishing between sequentially-processed and parallel-processed structures, with different evaluation formulas for each.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is separated into an external DB to improve operational efficiency and resource utilization, then scalability and resource efficiency are improved, but call processing Turn Around Time increases due to additional external DB access
Solution Approach 1:
The patent segments the software into multiple portions and classifies them as either sequentially-processed or parallel-processed structures. This segmentation allows for differentiated evaluation expressions to be applied to each type, enabling accurate performance impact assessment of data separation without actually implementing the separation yet.
Solution Approach 2:
The patent performs preliminary performance impact evaluation through simulation before actual data separation is implemented. By using log acquisition units to collect access patterns and evaluation units to calculate performance impacts on sequentially- and parallel-processed portions, the system predicts TAT extension in advance, allowing planners to make informed decisions about whether to proceed with data separation.
2Adaptability or versatility
If data separation is implemented to achieve cloud environment advantages, then system scalability is improved, but performance assessment becomes difficult due to complex software structure
Solution Approach 1:
The patent divides the complex software into sequentially-processed portions and parallel-processed portions, applying different evaluation expressions to each. This segmentation makes the performance assessment manageable despite software complexity, as each portion can be evaluated independently using appropriate formulas.
Solution Approach 2:
The patent uses log acquisition units to collect actual access patterns and feedback this information to the evaluation unit. This feedback loop enables the system to accurately assess performance impact based on real usage data rather than theoretical models, improving measurement precision for scalability decisions.
3Ease of operation
If individuality of processing servers is maintained, then processing flexibility is preserved, but resource efficiency decreases in cloud environment
Solution Approach 1:
The patent segments software functionality to identify which portions contribute to server individuality and which can be externalized. By evaluating the performance impact on sequentially- vs parallel-processed portions separately, the system can determine the optimal boundary for data separation that balances flexibility preservation with resource efficiency improvement.
Data Source
AI summary
To provide a performance impact evaluation apparatus and a performance impact evaluation method in which it is possible to more accurately evaluate a performance impact caused by software data/process separation. A performance impact evaluation apparatus 1 evaluates a performance impact caused by separating data 2 used in software into an external DB 4. The performance impact evaluation apparatus 1 includes a log acquisition unit 21 configured to acquire a log for recording an access to the external DB 4 and an evaluation unit 22 configured to evaluate the performance impact by using an evaluation expression considering a software structure in addition to characteristics of the data 2, based on the log acquired by the log acquisition unit 21.