How to use this monitoring glossary
Monitoring terms are only useful when the measurement boundary is clear. Use each definition to identify what is being measured, which context should travel with the signal and which operational decision the signal can support. For deeper implementation detail, follow the related guide links on each term page.
Start with the signal, then the workflow
Metrics summarize numeric behavior, logs preserve event detail and distributed traces follow request paths. Availability and latency describe service outcomes from different angles. OpenTelemetry provides standardized ways to generate and move several of those signals. The definitions are connected because real incident response depends on combining them rather than memorizing them separately.
Definitions should be operational
A useful definition makes it possible for two people to measure the same thing. That is why the glossary emphasizes boundaries, units, context and common failure modes rather than dictionary-style one-line descriptions alone.
Use the terms consistently across dashboards and incidents
Teams lose time when “latency,” “availability” or “uptime” mean different things in different reports. Link each important term to the exact measurement boundary, unit and data source used in your environment, then keep that convention stable in alerts, SLOs and post-incident reviews.