Ekta Khatana is technically sophisticated software testing engineer from Gurgaon and has seen Gurgaon grow up with her. She did her Bachelors in Information Technology from Banasthali University and has more than 7 years of experience working in Quality Engineering team of projects ranging from Banking domain to Event and Hospitality cloud. She has been associated with Cvent since past four years and currently working as Lead SDET. She is skilled in UI Automation tools like Selenium, WebDriver IO, Cucumber. Has gained expertise in ReadyAPI, Postman and Karate for API automation. Has been involved in performance testing using Gatling, Sitespeed and also been working on different monitoring tools like Datadog and Splunk. Apart from the technical stuff, she has a passion for adventure, be it trying something she has never done before like adventure and water sports or travelling and exploring different places. Her greatest attribute is having life-changing epiphanies every night before sleeping and forgetting about them by the morning.
Shift-Left Performance Engineering with CI/CD
Performance is a critical non-functional quality aspect in software development. Performance issues discovered at a later stage in development cycle can lead to more significant problems down the road, where resolution can be expensive and time-consuming. RECOMMENDED SOLUTION. Continuous integration and delivery coupled with automated performance tests will enable finding performance issues while code is being developed. It is a faster and cost-saving approach for releasing new product features and enhancements with minimal performance risks drastically reducing time and efforts. As we progress towards shifting left, it becomes crucial to ensure that we find functional and non-functional issues early in the release cycle. At Cvent, we have implemented an end-to-end automated process for continuous performance engineering, minimizing the monitoring time to check for any degradation. Before going to production, code is deployed to the pre-production region, where the load tests are executed and analyzed to check the performance bottleneck in nightly regressions. The process includes data setup at runtime using test Data Management, test execution, and report generation using Gatling. As a part of release certification, we have scheduled load regression test runs via Slackbot, which can be triggered with either Auto-Abort or Notify-Only functionality and thereafter a notification is sent. Before the test kicks off, the system checks if auto-scaling is enabled, then accordingly scale up the services depending on the service utilization metrics. When anyone executes the load tests from Jenkins Pipeline, results/metrics are collected just after the completion of load tests. A consolidated report detailing graphical data analysis and performance metrics for Gatling, Datadog, and Splunk is generated once the test is completed via a customized reporting framework.