1 research outputs found
Automated Database Selection Via Empirical Validation and Cost-Adjusted Scoring
No full textSelecting a database service can be challenging when decisions rely on static information or generalized benchmarks, which may not align with an application\u27s specific runtime performance needs, potentially leading to implementation risks. A system can provide automated, performance-validated database recommendations. The system can translate a user\u27s non-functional requirements into a profile used to programmatically provision temporary test environments for candidate databases. A workload generator can then simulate a user\u27s specific application load within these environments, which may allow the system to harvest real-time performance and cost metrics. The collected empirical data can be processed by a scoring algorithm to calculate a cost-adjusted performance score. This process can provide a data-driven justification for selecting a database configuration that is empirically validated to be compatible with a user\u27s performance and budgetary requirements
