Help to configure Collectd

[Patrick M. Hausen]

Haven't found the time yet. You need to learn Grafana and InfluxQL if you want to create dashboards.

If you want to use a complete dashboard you downloaded from some source you need to ask the author which time series database (Influx, Graphite, ...) they are using and how their template is set up.

Grafana is not a one stop solution.

[Dexter_23]

Ok but in my case what type of dashboard i need for my scenario?

InfluxDB o Graphite?

[Patrick M. Hausen]

Influx since you are using Influx as the time series database.

But then this depends on the template. Dashboards are in no way "standardised". What exactly do you want to achieve in the end?

Collectd just sends a series of entries to the time series database. They are structured like this:

servers.foo.bar.baz.quux <measurement> <timestamp>

The template in the influxdb config says that each line starts with the literal "servers" and that the following fields are named "hostname", "resource" and "instance" respectively.

You can then reference these fiels in Grafana like in my screenshot.

"Select the field value of gauge.system" from the default data source (InfluxDB) where the hostname matches the host selected by the user, the instance is named "cpu-average" and the resource is named "aggregation". You get these things from the "show series" output in Influx. That shows you all that is available.

Again, there is no one stop dashboard. If you name the field "foo" instead of "instance", you must use "foo" in Grafana - it's all sort of like "programming".

You probably might want to start some reading on Grafana. There's a Grafana forum, too.

[Dexter_23]

i don't have thies field like your screenshoot i think...