archibate System Monitor
Check current system health and report alarms when any resource exceeds safe thresholds.
Goal
Present a single summary table of CPU, memory, disk, and network status with OK / Watch / ALARM indicators. List top resource-consuming processes. Flag any metric that exceeds its threshold.
Thresholds
| Resource | Watch | ALARM |
|---|---|---|
| CPU | > 60% | > 80% |
| Memory | > 60% | > 80% |
| Disk | > 70% | > 85% |
Steps
1. Gather system metrics
Run the following commands and parse their output:
- CPU:
top -bn1 | head -5— extract user%, system%, idle%, and load averages - Memory:
free -h— extract total, used, available - Disk:
df -h --total -x tmpfs -x devtmpfs -x efivarfs— extract total, used, available, use% - Network:
cat /proc/net/dev— extract RX/TX bytes for non-loopback interfaces - Top processes:
ps aux --sort=-%cpu | head -6for CPU hogs,ps aux --sort=-%mem | head -6for memory hogs
2. Evaluate thresholds
Compare each metric against the threshold table above. Assign status:
- OK: below Watch threshold
- Watch: between Watch and ALARM thresholds
- ALARM: at or above ALARM threshold
3. Report
Present results as:
- A summary table with columns: Resource, Status, Details
- A Top Processes section showing the top 5 by CPU and top 5 by memory (deduplicate if overlapping)
- A Network section showing per-interface RX/TX rates
- If any metric is ALARM, lead with a bold warning line
Success criteria: User sees a clear, scannable health report and any alarms are immediately obvious.