Swe Intern
Atlanta, Georgia, United States
- Engineered a service that ensures the persistence of Kubernetes testing environments for the team's testing suite leveraging GitHub Actions, NodeJS, NextJS, Redis, Bash, and the Google Cloud Platform.- Deployed through Helm and Google's Kubernetes Engine for uptime and reliability.- Implemented automated microservice redeployment and retesting on the Kubernetes test cluster using Python.- Developed a dynamic diagnostics analyzer using Python, GitHub Actions, and Cypress.- Significantly improved debugging efficiency, reducing engineers' debugging time by 1-2 days per week for a 100+ member team, with the tool being successfully utilized for 3 weeks prior to the end of the internship.Working with the test automation team, we built a service that solved the problem of engineers not being able to access the Kubernetes cluster that the automated testing suite runs on. Previously to retest changes and fixes to a microservice, engineers would have to fully re-run the testing pipeline. This process involved downloading 20+ microservices, deploying a fresh test cluster, rebuilding the services, and re-executing the test which took around 20 minutes to fully complete. Our service streamlined this process, enabling engineers to extend the life of cluster environments. They can also clone environment replicas to their local machines, trigger automated rebuilding and redeployment of only the altered services on the cluster, and rerun the testing suite, completing the task in just 5 minutes.Furthermore, I developed a diagnostics analyzer capable of parsing through broker logs to identify deviations from the expected log content. This tool drastically improves the time in identifying microservices that generate unexpected responses or no responses at all.