Easy metrics dashboard for multiple systems?
from dataprolet@lemmy.dbzer0.com to selfhosted@lemmy.world on 01 Jun 2024 10:58
https://lemmy.dbzer0.com/post/21401398
from dataprolet@lemmy.dbzer0.com to selfhosted@lemmy.world on 01 Jun 2024 10:58
https://lemmy.dbzer0.com/post/21401398
I’m looking for a simply solution to monitor all my servers and systems using a single dashboard. I want to see metrics like CPU usage, used RAM and storage to see if something is wrong.
I just set up Node-Exporter, Prometheus and Grafana but haven’t found an existing dashboard that shows multiple hosts at once. Now I looked into Checkmk and Zabbix but I feel like both are a little overpowered for what I’m looking for.
Do you have any recommendations?
threaded - newest
Using the same setup, only have one host, but I remember there were some dashboards for multiple hosts, keep looking^^
I used the following dashboard now, but it’s not really satisfying and also doesn’t really fit more than 4 nodes. grafana.com/…/11756-hpc-node-exporter-server-metr…
I think you can use grafana to present vidgets from different dashboards in one.
PRTG maybe? It’s free for 100 sensors.
But it’s proprietary, unfortunately.
Sorry didn’t realize you were trying to connect the systems you mentioned, I thought this was a unified approach instead of them.
Cockpit is a dashboard (and control!) app that can connect to multiple hosts.
Thanks, looks promising. I’ll give it a try.
Question: why don’t you crate your own Grafana dashboard? Do you have too many servers?
I don’t want to configure a whole Dashboard for at least CPU, RAM, Storage and Network for up to 5 hosts.
You’re likely not going to find a premade dashboard that does exactly what you want, but grafana is extremely powerful if you’re willing to put in the time to learn it. There are ways to visualize things across hosts without having to configure things separately for every host. If you’re using the same mechanism to scrape metrics from each (sounds like you’re using prometheus + node exporter?), this could be as simple as adding a
by (node)(or whatever the label name is if it’s not node) grouping to the query on each panel.Exactly this. We use node exporter, Prometheus, and Grafana at my place of work to get node metrics across our K8s cluster for CPU, memory utilization, file system space, etc.
You’ll have to do some searching and tweaking of existing dashboards, but Grafana is crazy good
It’s not that bad. Grafana has a lot of features for this.
Btop running in multiple ssh sessions lol
Edit: lol you said “easy”
This is actually how I do things when working on remote machines. I have far too many monitors, so dedicating on of them to a handful of btop/nvtop terminals works pretty well.
I admit that it’s a less than perfect setup though, and a single program which could handle the remote connections internally and display an aggregate would be nice.
Check out Netdata
I use Homepage for something similar to this