<![CDATA[As the complexity and user volume of web applications continue to grow, performance testing has become increasingly crucial to ensure seamless operation and reliability. This necessitates effective testing and bottleneck identification strategies. This research focuses on designing a schematic mechanism for identifying potential performance bottlenecks under extreme conditions through a series of server performance tests involving direct observation of server application and database activities at varying CPU frequencies. Performance testing employs a load testing model to assess server behavior and response when handling simulated sequences of activities from virtual visitors simultaneously filling out online employment forms in a relatively short period with a large number of users. The primary tools utilized are Apache JMeter as a load generator equipped with the JMeter Stepping Thread plugin. Elasticsearch serves as the repository database for sensor data collected by Beats agents deployed on both application and database servers, either directly or through Logstash. Kibana, a graphical tool, interprets data retrieved via Kibana Query Language (KQL) from Elasticsearch into visual dashboard representations, which are further analyzed to identify potential and actual performance bottlenecks.]]>
Copyrights © 2024
