Set up the current OMD Labs Edition with Naemon and Thruk, brought in a bunch of config files from an old Nagios 3 setup, and it’s been an impressively smooth experience. Great job of packaging there.
Trying to see how Grafana’s graphs would look while keeping PNP4nagios too, I’ve followed instructions up to “Note you have to choose a primary graphing solution in thruk by using the right action_url in your hosts / services” (at https://labs.consol.de/de/omd/howtos/grafana/). Assuming that’s to be in /opt/omd/sites/[sitename]/etc/naemon/conf.d, I’ve tried altering pnp4nagios.cfg (as well as check_mk_templates.cfg, which didn’t seem to accomplish anything), substituting an action_url for Grafana based on an old post here. At this point I’ve getting some (basically empty – just 3 horizontal lines representing thresholds) Grafana screens for hosts, but the only graphs that come up in any context for services are still RDD. I did run the provided script to copy PNP4nagios data over to Grafana, and set up Gearman. Maybe that’s not compatible with turning on InfluxDB, which I also did, not being sure which section of the instructions fit with each different scenario (that is, two graphing engines on, compared with just one or the other).
Which leads to two questions: Where might I find fuller instructions for this setup? What does Grafana bring to the table here, when properly set up? Obviously it’s a rich solution in other contexts, but is it worthwhile in OMD?