We have problem tuning the memory management of JVM\'s. The very same application running on the k8s cluster, but one of the pods\' jvm heap usage rises to ~95% and, when we try