For now, add only one metric, a counter for the errors occurring while scanning pod resources on the node.