Read Quick View and timed charts
Use Quick View to decide which resource deserves a closer look, then use chart time to qualify what you saw. This is an inspection task for an installed, running Fleet Console on macOS 14 or later; no service-control permission is needed.
Compare a glance with history
- Click the menu-bar cluster to open Quick View. Read CPU and memory separately from disk capacity and directional network rates. A full disk and a busy CPU describe different constraints.
- Note the resource and approximate observation time. Do not copy a staged illustration's values into an incident report.
- Open the dashboard with Command 0 and use Overview to compare the histories with the largest process consumers.
- Focus CPU or Memory. Move between 15 seconds, 30 seconds, 1 minute and 2 minutes to see whether the high reading persists across the observed window.
- Open Processes with Command 2 if you need a current process to inspect. Use Command K in that destination to narrow the list.
Check the interpretation
A useful finding names both the resource and its time context: a short spike in the observed interval, sustained activity, or insufficient observations. The horizontal axis uses observation times, including gaps. Widening the range does not create measurements for time the app did not sample.
Quick View and the dashboard use the same probes; they are not independent confirmation that a service works. Use service health evidence for that different question.
When the views do not answer the question
If a sensor is unavailable, retain that label. If the time window contains a gap, describe it instead of joining the surrounding values into a continuous measurement. If a process is no longer present, do not infer its current load from its previous row. Return to the live process list and inspect a current identity.
The maintained Fleet Console overview describes these chart and sampling boundaries. See sampling reference for how to phrase missing evidence.