What do y’all use to monitor many linux servers?

[regalpotoo@lemmy.world]

• Base ansible role installs Prometheus node exporter, configured with the text file collector
• VM automations push DNS records so that the Prometheus dns-sd automatically discovers them
• Ansible roles for add Cron jobs that generate metrics for specific systems and dump them for the text file collector
• Grafana for dashboards
• Karma as a UI in front of Prometheus alert manager

[laintrain@lemmy.dbzer0.com]

Cockpit.

[hobbsc@lemmy.sdf.org]

is cockpit on a server by server basis or can you monitor multiple servers with it?

[andromxda@lemmy.dbzer0.com]

Netdata is exactly what you're looking for. It's basically an all in one monitoring and and alerting suite that collects and analyzes data, and provides a gorgeous web dashboard for you to view.

You can also manually replicate this using Prometheus, Grafana and other tools, but that requires a much bigger effort to set up.

Edit: There's a public demo instance where you can try everything out: https://frankfurt.netdata.rocks/

[ikidd@lemmy.world]

I think they went to 5 nodes max on the free version as of the last patch. That's damn near useless.

[darkdarkhouse@lemmy.sdf.org]

I use my family. It has a simple volume based alert for when services are offline.

[vfsh@lemmy.blahaj.zone]

It'll even automatically configured variable alert volumes corresponding to the importance of the service!

[toribor@corndog.social]

The five node limit is a dealbreaker for me too. I'm also annoyed the free version doesn't have any real built in options to secure data by default. I followed a TechnoTim tutorial to get the NetData/Prometheus/Grafana stuff setup but it was too limited and required too much manual effort.

[ipkpjersi@lemmy.ml]

Seconding Netdata, I've been using it for years. It's pretty great.

[ipkpjersi@lemmy.ml]

Is that just for the centralized dashboard portion? I tend to use each instance of it standalone, and primarily for the email alerts.

[toribor@corndog.social]

Any chance you'd be willing to share playbooks or point me toward any resources you used?

I use Ansible to manage config across all my workstations/servers but I haven't gotten around to automating log shipping yet or aggregating system metrics.

[ikidd@lemmy.world]

I believe so. I imagine the next stage of the enshittification will be to force those standalones to register with a portal account.

[ipkpjersi@lemmy.ml]

That would be a truly dark day. I never liked their centralized dashboard functionality, it always seemed cumbersome to me.

I hope that doesn't happen, but I guess if it does, I will really need to find a different monitoring tool.

[cmc@discuss.tchncs.de]

You can monitor multiple machines via the host switcher menu at the top-left of the screen: Multiple Machines

[andromxda@lemmy.dbzer0.com]

Oh that sucks. I haven't used it personally in quite a while, since I switched to the Grafana stack

[fmstrat@lemmy.nowsci.com]

Until the UPS battery gets low and it beeps, and they look for a way to turn it off vs calling you. Yup.