I tried to increase load for my pod in Open shift by doing load testing, I was doing a lot of parallel request for my application and I see that my pod starts to use much more cpu and memory and it's completely normal, but problem is that even next day I see that my pod memory usage is still as high as it was, even no one is doing any request to that application. Does Open shift pod not able decrease memory usage for pod if it sees that usage decreased or I need to write specific command in config to be able to do that?
Problem was that app which we host in Openshift, had a memory leak which increased cpu usage overtime till pod reached cpu limit and restarted itself. If you see that your pod memory usage never decrease, you should check if your app do not have memory leak, because in normal circumstances memory usage should decrease.