This study examines the scalability and stability of resource allocation in a containerized cloud environment, focusing on how different configurations of Virtual Machines (VMs) per Physical Machine (PM) handle increasing task arrival rates. By simulating different disposition of the proposed architecture, we assess the performance and resource utilization under varying workloads. The findings indicate that the 40 VM configuration offers the best performance, exhibiting a linear increase in utilization while avoiding saturation and maintaining efficient resource usage. In contrast, increasing the number of virtual machines, although stable, shows significant underutilization, indicating inefficiency. Configurations with less VMs reach saturation early, highlighting potential instability under higher workloads.