Server Metrics

hav.sh can collect and display real-time server metrics using Telegraf. Once enabled, you get charts for CPU, memory, disk usage, disk I/O, network, system load, and Docker container metrics — directly in the dashboard.

#Enabling server metrics

#1. Open the server settings

Navigate to Servers, select a server, and go to the Configuration page.

#2. Enable Telegraf

In the Monitoring section, switch to the Telegraf tab and toggle it on.

[!NOTE] Telegraf cannot be disabled once enabled. A lightweight agent is installed on your server that collects metrics every 5 minutes.

#3. Wait for first data

After enabling, the Telegraf agent will be deployed automatically. The first metrics typically appear within 5 minutes.

#Viewing metrics

Navigate to Servers → select a server → Metrics. You'll see charts for:

• CPU Usage — Percentage of CPU in use
• Memory Usage — Used memory and cache/buffers as a percentage of total
• Disk Usage — Percentage of disk space in use on the root partition
• Disk I/O — Bytes read and written per interval
• Network I/O — Bytes received and transmitted per interval
• System Load Average — 1-minute, 5-minute, and 15-minute load averages

When Docker containers are running on the server, a Docker Containers section appears with:

• Container CPU — CPU usage percentage per container
• Container Memory — Memory usage in bytes per container
• Container Network I/O — Combined network traffic (RX + TX) per container per interval

When a PostgreSQL database is deployed on the server, a PostgreSQL section appears with:

• Connections — Current connection count with max connections reference line
• Database Size — Total size of all databases
• Transactions — Commits and rollbacks per interval
• Cache Hit Ratio — Percentage of reads served from buffer cache (healthy is above 99%)

#Time range

Use the time range buttons at the top of the page to switch between:

Charts auto-refresh every 60 seconds.

#Dashboard overview

When at least one server has Telegraf enabled, the main dashboard shows a Server Metrics widget with:

• Avg CPU — Average CPU usage across all monitored servers
• Avg Memory — Average memory usage across all monitored servers
• Highest Disk — The server with the highest disk usage
• Monitored — Total number of servers with Telegraf enabled

#Data retention

Metrics older than 30 days are automatically pruned to keep your database lean.