Monitor Your System with Grafana using Netdata and Prometheus
HTML-код
- Опубликовано: 27 авг 2024
- Prometheus is quite amazing! Here I show you how to use it to pull metrics from the Netdata API to display in a custom dashboard with Grafana.
Written guide:
thehomelab.wik...
Tags used in this video
grafana dashboard, grafana, prometheus, netdata, homelab, self hosting, system monitor, server monitor, open source, geekedtv, grafana tutorial
![Oasis - Oasis Live '25 [Official Trailer]](http://i.ytimg.com/vi/gol_JnuEtKM/mqdefault.jpg)
Was this helpful? I hope so! Let me know if you set this up yourself and if you have any issues.
Nice, I've been trying to set up something similar to this myself and found this super useful, thanks.
Hey! Do you maybe have the Grafana Dashboard json available somewhere?
Thanks for the good tutorial. Have mine setup and monitoring four servers in my proxmox cluster.
Thanks for watching Matt!
Thank you for a clear and concise video. I just completed this. For anyone running all this on a Synology NAS, I had to add "free" and "cached" queries in Grafana to see my total free ram. You can add queries together in the Transform tab.
It's odd how Netdata seems to be very accurate with RAM available but used is an issue I ran into too. I took a look at cached but it too was off by a fair amount compared to CLI tools when I compared it.
@@selfhosted That is odd as "used" appears accurate to me when looking at my Synology dashboard. I switched to using netdata_mem_available_MiB_average instead of netdata_system_ram_MiB_average and think I'm getting a more accurate result for available RAM
@@gnote2greatness Right. This metric is for available memory, not used. Used is what is not so accurate for me. What metric are you using for used ram?
@@selfhosted Sorry, RAM used from your video was closely accurate for me. 3.63GB using top and 3.81GB using netdata. Since you're RAM available is accurate, you could take your total ram and subtract netdata_mem_available_MiB_average. For me it was 15483 - netdata_mem_available_MiB_average. Not the greatest solution but might get you closer.
@@gnote2greatness Interesting concept... but it works much better. Thanks for the tip!
clear to the point, and thanks to written guide. Gained a sub
You're so welcome!
bro thank you very much for the guide with a simple explanation of the necessary things⬆
Yes!!! Finally!!!
I'm always nervous with long videos like this because the editing takes so long and I don't want to feel like I rushed through it. So let me know if this is good enough. Thanks for watching as always!
why use netdata in this scenario? why not send data directly to prometheus?
Through what, a node exporter?
@@GrishTech did you watch the video? 4:54 prometheus is already scraping the target
Great Video
Thanks Craig!
Can you post the JSON for you Grafana dashboard? The link in the written guide is dead. Great vid, just what I needed.
1:28 I am so confused about the transition from proxmox to portainer. Are you installing prometheus and grafana as an LXC container or Docker container managed in Portainer. Please clarify as that literally made no sense to me.
Late to the party, but: LXC container with Docker installed. Running Portainer on on that LXC via docker run. Then setup grafana/prometheus as Portainer stacks. That's all there is to it.
Thanks for the video I was able to install this and monitor my raspberry pi. Just a question though how do you find out what metric to use on Prometheus? is there a documentation for it? I'. tying to find the disk read/write/utilizations.
It's in the Prometheus explorer. You can see all the metrics there.
Nice video! Can you make another video where instead of Prometheus scraping the Netdata hosts, Netdata agents send data to Prometheus? (So a push instead of a pull) Thanks!
omg thanks
Wow, Just wow
Pretty awesome isn’t it?
@@selfhosted totally, the way you describe everything, awesome! I never used netdata and prometheus before but from this video I learned many things
Do you know of any IT/pc-parts inventory web interface application ????
snipeitapp.com/
great video , but looks like the link dont work , several youtube channels has this link problem
your website is down
Shit, too much information. Simply i just remove my freshly created grafana folder and stay with synology dsm manager.
I really enjoy your presentation style, thank you.
Since Portainer only supports Compose V2 file format how can you install your stack which is V3.3
Thank you! What version of Portainer are you running? Mine worked fine as you saw. Try changing it in the stack and see if it works for you.
Michael maybe a version issue. I am running portainer 2.6.3, and following the video I have been able to use version 3.3 for the stack creation.
@@turtleneckzor I believe version Portainer 2.6.0 upward supports 3.3.