if a device reports log a message such as "Device CPU Utilization has
exceeded a threshold", then XDAS_AE_RESOURCE_EXHAUST seems to be the proper
taxonomy. But what would be the correct outcome? XDAS_OUT_SUCCESS because
the performance sensor worked correctly. Or XDAS_OUT_THRESHOLD_EXCEEDED? But
that is a detailed failure reason. So the action of exhausting a resource
failed?? And which outcome would be used if the CPU utilization falls below
the threshold again?

All this seems quite counterintuitive. I think we need some extensions or at
least a clarification to our taxonomy because state notifications like these
don't fit well into the framework XDAS provides.