We are using Puma with cluster mode where we are using 3 workers with 5 threads. Once the container application is up it starts consuming memory by Puma workers at a high rate and after some time memory utilization by workers slows down but increases step by step. After the memory is utilized 100% it stops the container and starts again.
This is really a big problem we are facing.
Attaching the Memory utilization Graph for Rails Puma Application.
