Uptime depends on how availability is measured during the particular period. For evenly spaced checks, the rate of successful checks provides an approximation.
Step by step
-
First determine the monitoring period and control interval. For example, query external Minecraft status at the same frequency for a week.
-
Record each sample as pass, fail, or no measurement. If you want timeout and planned maintenance, show them separately in the report.
-
Calculate the coverage rate and success rate together. If 90 out of 100 expected samples are measured and 81 are successful, measured success is 90% and coverage is 90%.
-
Write the time the data was last updated and the point at which it was measured. Compare the result with the input test from different network when necessary.
Keep in mind
History records that are kept subject to change are not the same as regular time instances. Extracting exact running time from this data.