Currently the operator's csv does not specify memory and cpu requests/limits. The action is to investigate whether default values should be included in the csv versus limit ranges, etc. As the collector's resource requirements will track traffic through the application network, more detailed sizing recommendation and associated documentation will likely be needed.