International Journal of Scientific Research and Engineering Development

Because of the scalability, portability and rapid deployment, Containerized applications are widely adopted in cloud environments. With Docker's Restart policies and other automated self-healing mechanisms, the modern container platforms provide recovery from failures. However the real recovery experience from the user's perspective is not interpreted by traditional evaluation metrics like uptime and restart frequency. This paper presents a new metric named as Recovery-Complete Response Delay (RCRD), which evaluates the time period that passes when the fault occurs till the successful responses of restoration of the application. Using AWS EC2 instance, a Dockerized NGINX container was set up with automatic restart policies in order to analyze the efficiency of RCRD. Faults like out-of-memory (OOM), CPU stress and forced process termination were introduced intentionally. Automated monitoring scripts were used to solidify temporal accuracy. The findings showcases the limitations of traditional metrics and points up the importance of time-based evaluation. RCRD provides user-centric and thorough measure of system resilience in cloud-native applications.

Aditya Deo, Kalpesh Gaikwad,"Measuring Self-Healing Efficiency in Docker Containers Using Recovery-Complete Response Delay on AWS EC2" International Journal of Scientific Research and Engineering Development, V9(2): Page(1586-1590) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.