Repository navigation
Metrics and Status
Alertivity normally samples about every five seconds. CPU and throughput readings need successive samples to calculate activity, so initial readings can be zero or show that data is being collected.
The overall status follows the highest severity among CPU, memory, disk, and network. Normal means all readings are below their elevated thresholds; Elevated and Critical mean at least one reading reached that level. At equal severity, trigger priority is CPU, memory, disk, then network. Critical means a threshold was reached. It does not indicate a hardware fault.
| Metric | Elevated at | Critical at | Measurement |
|---|---|---|---|
| CPU | 50% | 80% | Total system processor utilization |
| Memory | 70% | 85% | Used physical memory as a share of total |
| Disk | 30 MB/s | 120 MB/s | Combined read and write throughput |
| Network | 5 MB/s | 20 MB/s | Combined receive and send throughput |
Values below Elevated are Normal. At a threshold, the higher status applies: CPU at 80% is Critical with the defaults. MB/s uses decimal megabytes (1 MB = 1,000,000 bytes), not megabits. Customize these values under Settings → Thresholds with Pro in the App Store build, or without a purchase in the open-source build; process detection has separate controls.
A summary such as “Alert thresholds reached: CPU 80% and Memory 70%” shows the configured thresholds reached by current activity. The live readings are in the metric rows below. “Reached” includes equality; thresholds flag activity and do not cap resource use.
Each affected metric shows the Elevated or Critical threshold for its current severity, including custom values. Critical metrics appear first; CPU, memory, disk, then network break ties.
Memory usage is calculated from physical memory minus free and inactive pages. It is not the same measurement as Activity Monitor’s Memory Pressure graph. Network totals exclude loopback and several virtual/tunnel interface types to avoid double counting; they are not a per-app traffic meter or a connection speed test. Disk activity is throughput, not remaining storage space.
If a system reading is unavailable, Alertivity may keep its last value. It also retains the previous disk reading when a sample reports zero. If a reading seems stuck, compare it with Activity Monitor.
Auto switch chooses a Critical metric, using the order saved in Settings → Menu Bar → Metric priority (CPU, memory, disk, then network by default), after two consecutive matching samples. When none is Critical, it returns to the chosen default after the same sampling check. Selection uses severity and your metric priority, not the raw values. The custom order affects only Auto switch.
See Settings and Notifications and permissions.