A workaround, a configuration record and a measured commitment
A fictional service agreement counts every minute in a ten-thousand-minute observation window and has no maintenance exclusion in this exercise. The service is unavailable for sixty minutes. An operator restores access by temporarily routing requests to a verified spare instance, but the triggering defect remains under investigation. The configuration record still names the failed instance as the only dependency, and the temporary route has no assigned expiry owner. A status message claims that restoration has proved the problem permanently resolved and that the stale configuration record can be deleted. The team must distinguish incident restoration, problem investigation, configuration accuracy and controlled follow-up. No live change is requested by the scenario, and the availability formula applies only to the supplied contractual measurement model.
1. Calculate measured availability and identify the incident/problem distinction.
Read the worked answer
Availability is(10,000−60)/10,000=99.4 percent. Restoring acceptable operation addresses the incident, while investigation and treatment of the underlying cause can continue as problem work.
2. What follow-up is needed for the temporary route?
Read the worked answer
Record the actual dependency and change, assign a responsible owner and expiry/review condition, verify monitoring and recovery implications, and investigate the cause. Deleting the record would conceal useful operating evidence rather than resolve the defect.