At a random moment one of the ARR devices started to fail. The cause was found in a part of the configuration at the virtualisation and networking level. Following the incident all devices were implemented with the latest best practice configuration after which the issues no longer occured.