+1 for Uptime Kuma. I use it in conjunction with a tiny Go endpoint that exposes memory, disk and cpu. And, like @iii I use ntfy for notifications. I went down the Grafana/Influx etc route - and had a heap of fun making a dashboard, but then never looked at it. With my two Kuma instances (one on a VPS and one in my homelab) in browser tabs, and ntfy for notifications on my watch, I feel confidently across the regular things that can go wrong.