Centralized Performance Monitoring
Cockpit acts as a central control room that monitors the CPU, memory, and network health of all your physical servers and virtual machines, showing their performance in visual graphs.
Metric Collection & Telemetry 📈
💡 Analogy: The Hospital Vitals Monitor
Think of metric telemetry collection like a hospital patient monitor connected to sensors:
- The Sensors (Telemetry Source): Cockpit's collector service acts like finger clips and heart rate sensors attached to each Vapor server, reading their live performance data.
- The Pulse Check (Interval): The system checks the vitals of every server and virtual machine every 10 seconds to make sure they are running smoothly.
- Downsampling & Storage (Retention): To prevent your storage drive from filling up with millions of tiny logs, Cockpit keeps second-by-second data for a short time, then compresses it into hourly and daily averages (called downsampling). This allows you to view historical graphs for weeks or months without slowing down the database.
Centralized Dashboard Views
Cockpit organizes this performance data into three easy-to-read dashboard levels:
1. Cluster Overview Dashboard (The Whole Hospital Ward)
Gives you an aggregate view of your entire server infrastructure:
- Total Resource Usage: Shows the total number of CPU cores and RAM in your cluster, and what percentage is currently active.
- Storage Capacity: Lists how much storage space is left across all your disks.
- Server Health Status: A quick list showing which physical servers are healthy (
connected), disconnected, or experiencing issues.
2. Host Detail Dashboard (Single Patient Chart)
Displays performance stats for a single physical server:
- Live Charts: Real-time graphs showing CPU, Memory, and Disk read/write speeds.
- Hardware Profile: Lists basic hardware details (like the manufacturer, BIOS version, operating system kernel, and how long the server has been running since its last reboot).
3. VM Sparklines & Detail Metrics (Individual Patient Vitals)
- VM List Sparklines:
- What is a Sparkline? It is a mini line graph displayed directly next to each VM's name in your lists. It acts like a tiny heartbeat line, showing at a glance whether a VM is idling or working hard, without needing to open its detail page.
- VM Details: Shows historical performance trends, disk read/write bandwidth, and network packet rates for a single virtual machine.