Simple Homelab Health Monitoring with Discord Webhooks

I wanted a (VERY) lightweight way to keep tabs on my homelab server. I didn’t want to ssh to it periodically or a full monitoring stack like Prometheus/Grafana/Netdata. Here’s my solution:

The Setup

Two components make the magic happen:

1. Collector - Runs every hour via cron, appending CPU, memory, load, and disk usage to a CSV file. Data is retained for 7 days (I might have a hoarding issue).

2. Reporter - Runs daily at 9 AM, calculates min/avg/max from the collected samples, checks for journal errors and SSD health, then sends a formatted embed to a Discord webhook. I start work at 9 AM so this is another pleasant addition to the daily ritual email-checking and groaning over a cup of steaming tea.

What Gets Monitored

• CPU usage (min/avg/max over 24h)
• Memory usage (min/avg/max over 24h)
• Load average (min/avg/max over 24h)
• Disk usage (current)
• SSD SMART health status
• System errors from journalctl

The report color-codes itself: green for OK, yellow for warnings (high usage or errors present), red for critical issues.

Why This Approach?

• No dependencies - Just bash, cron, and curl
• No services to manage - Cron handles scheduling
• Minimal resource usage - A few KB of CSV data per day
• Discord integration - Notifications where I already am
• Actionable - Only alerts on actual issues with --errors-only mode

The title image to this post is the actual summary with all the stats at a glance. If there are errors in the journal, they’re included in a code block at the bottom. (The image is a test message from a snapshot which is why it’s got weird values. I will have to come back and update it but I am 100% certain I’ll never do that)

Source

The scripts are available at git.jawhng.xyz/jawhng/homelab-monitoring-webhook.