Setup Grafana and Prometheus for monitoring of servers

[robe]

When I killed off the old osgeo3, I made all the munin-nodes no longer useful.

I started to setup a new munin -- but couldn't get it to work -- which got me thinking maybe there is something better anyway.

Long story short -- after accidentally stumbling on some videos about Matrix, I came across a combination of using Prometheus with Grafana and it ended up surprisingly easy to install.

I've only installed nodes so far on the newer servers on osgeo3 and started to do it on osgeo7.

The web interface for the monitoring is at

​https://monitor.osgeo.org

and authentication is via our LDAP.

I'm going to try installing a node exporter on tracsvn next.

Attached is a sample chart in Grafana for nexus container

[robe]

sorry type in url - it's ​https://monitor.osgeo.org

[robe]

okay have configured tracsvn. A bit different than the newer OS because it's not systemd based.

The same binaries work since its all statically linked go.

I have the steps and binaries in this repo. I'll add the systemd steps in a bit.

​https://git.osgeo.org/gitea/sac/prometheus-config

[strk]

Great to see modern monitoring coming up ! Does grafana support alerting system ? Since you mention matrix... could grafana send alerts via matrix to #sac:osgeo.org ?

[robe]

move open times to new contract

[robe]

oops moved to wrong milestone last run

[robe]

Milestone renamed