In production 2.0 we are seeing that REDIS is dying every week or so. In the /var/log/messages there is a note that it is killed by the oom-killer (out of memory killer). I presume that this is because it is using all the memory that is allowed to it by the podman container in which it runs, because the full machine is never out of memory