OMV uses collecd to get system metrics.
collectd seems to be capable to write to prometheus.
For some time I had an influxdb running on RPi 1 and 4, so I know that the DB works 
If the idea "bears fruit", that would be a really great solution to visualize the measurements taken for very specific analysis of OMV servers.
When you have some early prototype, I'd be eager to see it
As a real world example I'm looking for a better visualization of SMB benchmark results (link in my signature below)
found it in https://github.com/collectd/collectd#readme
"- write_prometheus
Publish values using an embedded HTTP server, in a format compatible with Prometheus' collectd_exporter."
If you get it to fly, please please please write up something for us dummies to follow
Omvben I stumbled over a blog post covering the topic in depth as it looks. Did you know it?
did anyone get it working?
I have OMV5 with ioBroker, InfluxDB and Grafana deployed via Docker (Portainer)
Would be nice to send all system metrics either via iobroker to InfluxDB or directly to the DB.
not me
I'd also also like to try this out...
How much resources does combo Grafana+Node exporter+Prometheus use? I would like to deploy and use them, but if it will be something that will 24/7 use my CPU and RAM on some higher level, then I would pass.
I'm using netdata, but I would like to try grafana - prometheus also
Me too. But don't want it to be constant cpu and ram hog.
hi,
I have been playing around with said docker combinations. Have a look:
I did encounter a unresolved issue which I am still investigating.
Regaring the ram+cpu hog question, on recent hardware that's not something to worry about. You can look at the attached pictures in my post. I'm running on a j4105 quadcore with 16Gb dual channel.
These are the steps I took to get the collectd prometheus exporter up and running - Note that these are messing with your OMV install - I take NO RESPONSIBILITY! ZERO! NADA! I AM NOT RESPONSIBLE if this goofs something up on your system!
This has only been tried on my own system running OMV Version 6.3.4-1 (Shaitan) Linux 6.0.0-0.deb11.6-amd64 and I havent run it through any upgrades so I dont have any idea if it sticks or not. I also don't know if this is the best way to wire this up, I just know it works (seems like I should be able to use the existing web server, but I'm not sure).
First, you need to install two dependencies; The Collectd Core Utils which includes the write-prometheus plugin, and the tiny webserver library that helps serve up the data.
sudo apt install collectd-utils libmicrohttpd12Then you need to create a configuration file for the Collectd daemon to tell it to load the write-prometheus plugin. It should be a new file; /etc/collectd/collectd.conf.d/prometheus.conf and the contents should be:
# Configure a prometheus endpoint
LoadPlugin write_prometheus
<Plugin "write_prometheus">
Port "9103"
</Plugin>Be sure to create the file as root! You can change the Port to be whatever port you want to show the Prometheus data on, just be careful it doesn't conflict with any existing ports (including those of Docker Containers on your OMV host!)
Then, restart the Collectd application:
sudo systemctl restart collectdIf everything went well, you shouldnt get any messages. You can then go to http://<your_omv_host>:<port> and you should get a nice wall of text.
If something went wrong, you'll see an error message like this:
Job for collectd.service failed because the control process exited with error code.
See "systemctl status collectd.service" and "journalctl -xe" for details.
Follow those instructions and fire up your Google engines...
Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!