Worked on reliability and observability enhancements for the EnterpriseDB/cloudnative-pg project, focusing on the Kubectl cnpg logs plugin. Addressed a memory leak that caused unbounded memory growth during long-running log streaming by reusing a single timer across iterations, which stabilized memory usage and reduced the risk of out-of-memory errors in production environments. Utilized Go for plugin development, leveraging Ginkgo and Gomega for testing to ensure robust functionality. This targeted bug fix improved the stability of log-follow operations on busy Kubernetes clusters, enhancing the troubleshooting experience and strengthening cluster observability without introducing new features during the development period.