Image-Processing Workloads and DDoS Attack Resilience: Evaluating Docker and Podman Containers on Raspberry Pi and ODROID
Eric Gamess, Mausam Parajuli · 2024
Containerization has revolutionized software development and deployment. This study investigates the performance of Docker and Podman containers on various ARM-based Single-Board Computers (SBCs), including the ODROID-XU4, ODROID-N2+, and Raspberry Pi 4 Model B (RPi 4B), for image processing tasks, in addition to exploring the impact of Distributed Denial of Service (DDoS) attacks on these container technologies. The research assesses parameters such as time, CPU utilization, and memory utilization during image processing tasks. Moreover, the paper inspects the resilience of web servers under DDoS attacks when varying the retrieved file sizes, frequency of malignant traffic generated by the attackers, and web server choices, emphasizing the need to implement adaptive security measures. The ODROID-N2+ with 4 GB memory emerged as a standout performer, navigating the challenges posed by DDoS attacks while maintaining acceptable speeds, suggesting its potential as a cost-effective Docker and Podman engine. At the level of the container technologies, Docker and Podman performed similarly. Overall, the insights gained from this research offer valuable guidance for practitioners who are planning containerized deployments on resource-constrained devices.