Show HN: Lightweight system monitor for Linux VPS written in Go | Hacker News Reader

Show HN: Lightweight system monitor for Linux VPS written in Go

github.com

63 pointsygagaga19 comments

Cool. I went another way for monitoring my servers and built a local macOS app instead of installing the monitoring dashboard on the server itself.

https://kitaso.app/

It monitors any server you add to it and it can install a daemon that collects metrics for when you are offline. I keep it running to monitor our servers.

Wanted to quickly try this but I couldn't get password authentication to work. No password field appears, and I can't save the server entry since it can't authenticate.

If anybody is looking for alternatives, I’m a very happy user of Bezel (https://beszel.dev) on my home lab servers including a Raspberry Pi Kubernetes cluster and TrueNAS. Highly recommend.

Same, built on pocketbase too. And rather mature with strong docs now.

cool man, nice. Question, if I'm already doing alloy/tempo/prometheus/grafana does this give me something I'm missing? I don't mean that in a condescending way that a lot of people in forums say like "why would you build this when xyz already exists" as if "for fun" and "to learn" aren't good enough reasons haha. I'm just curious before I dig too deeply into it because I'm pretty busy if there is anything of note that lets this do something that the prometheus-metrics/grafana stack doesn't or doesn't do well. Thanks man!

A lightweight system monitor written in Go is an interesting project.

It also seems like a good fit for something I've wanted to try for a while: building the UI as a SPA, embedding the generated assets into the Go binary with `embed.FS`, and shipping everything as a single self-contained executable.

How is this better than Prometheus?

Not better - different purpose. Its super lightweight, easy to install, simple for one server

If I were to bet this is more a learning experience rather than trying to beat a product. Even better that this was done without AI (by the looks of it).

I’ve also done something similar in Go too, and several people used it, mostly because it’s easy to install/use and is lightweight.

https://github.com/joaomagfreitas/zero-monitor

Cool project! I like design of the dashboards a lot. Will try it out

It's a single git repository that compiles to a single binary that does really basic monitoring without any dicking about.

It lacks about 99% of the things that Prometheus can do, which is probably what you want for simple up/down healthy/not-healthy indication on a single box.

You start with not installing Prometheus! Prom. is a full on stats grabber and display system with knobs on.

This is a single system effort and at 5MB RAM is rather tiddly and surely counts as minimal overhead. It could probably send the metrics to Prom. or make them available for scraping with minimal extra effort but I think that should be left as an exercise for the interested reader.

The thing about gathering metrics is you have to be careful that the gathering/processing system doesn't drown out the thing you are measuring!

"curl -sL https://raw.githubusercontent.com/leodeim/vpsmon/main/instal... | sudo bash"

lol nope. i thought we all grew out of this crap?

just check the code before running :)

"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."

https://news.ycombinator.com/newsguidelines.html

https://tferdinand.net/en/why-curl-bash-is-a-dangerous-bad-h...

https://www.lesinskis.com/dont-pipe-curl-into-bash.html

https://www.cyberhub.blog/article/25147-why-curl-pipe-bash-i...

oh hey lets not forget this one too https://news.ycombinator.com/item?id=47860669

That's nice but you can also just look at the linked github repo and inspect install.sh for yourself. It's pretty straightforward. It builds the binary for your monitoring software from the git repo, based on the architecture you're on, and if required sets up a systemd unit to kick it off on boot.

It's quite nice.

Yes, I know that people repeat this a lot, but it's a generic dismissal that has nothing to do with the specifics of the project being discussed, which makes it shallow and offtopic. (Especially when taking the form of a snarky attack.)

Perhaps the first 1,000 times it was mentioned, there was some juice in the lemon, but it's long gone.

Did we? But my concern is with uninstallation:

"curl -sL https://... | sudo bash"

Now you won't be able to uninstall it if the repo ever goes black...

Reply on news.ycombinator.com