Under normal circumstances, a reasonable monitoring system will not
consume enough CPU and memory to be noticed. However, you should
test for failure modes. What happens if the monitoring server1 is down
or nonfunctional? What happens if there is a network outage? Also be
careful of transitions from one monitoring system to another: Remember
to turn off the old service when the new one is fully operational.