← BackProject · 09

VPS Monitor

Built after running a side project server blind for too long: a process leaked memory over several days and the first sign of a problem was the site going down. Needed something that could be dropped onto any Linux or macOS machine instantly, with no agent to install and no dashboard to configure.
Polls CPU utilization and virtual memory at a configurable interval and prints timestamped reports to stdout with ANSI color codes: green for healthy, yellow for approaching threshold, red for critical. Readable at a glance.
Thresholds are configurable at startup because a fixed alert percentage is meaningless across machines. A 512MB VPS and a 32GB dev box have completely different baselines for what 'high' means.
Single static binary. No daemon, no config files required to get started. Cross-compiles to Linux arm64 for cheap cloud instances; scp the binary and run it. That's the entire deployment.
Designed to be the first thing you check when something feels wrong, before reaching for htop or pulling up a full monitoring stack.

Tech Stack