What port do Prometheus listen on, for Grafana to connect?

[stoccafisso]

Probably stupid question. What port do Prometheus listen on, for Grafana to connect?

I tried http://prometheus.cloudron.tld:9090 (cloudron installation) but that did not work.

[stoccafisso]

I get this error message in Grafana:

Post "http://prometheus.cloudron.tld:9090/api/v1/query": dial tcp xxx.xxx.193.64:9090: i/o timeout - There was an error returned querying the Prometheus API.

[joseph]

@stoccafisso just the https port (all apps on cloudron are behind a reverse proxy). so https://prometheus.cloudron.tld

[stoccafisso]

@joseph I tried both with or without https://

[james]

Installed on the demo server, Prometheus https://pm.demo.cloudron.io and Grafana https://grafa.demo.cloudron.io

Added Prometheus as a new data source and used basic auth.

Save and test:

You can check it out https://my.demo.cloudron.io/ cloudron:cloudron as the login and Grafana default admin is admin:admin.

[stoccafisso]

Hi @james

Now I see, it was just to skip the port 9090 behind URL. I didn't even consider it possible. Thank you, it worked now

[james]

@stoccafisso Happy to help.

[stoccafisso]

But now I experience another problem. Grafana doesnt seem to condsider the new modified prometheus config file. Is there some cache stored in Grafana, or Prometheus, that I need to clear out in order for those two to see the new config file?

[stoccafisso]

Ahh...in Prometheus, just click "Status - Service discovery"

[stoccafisso]

You may set it to "solved"

[stoccafisso]

Hmm....Grafana seems not to pick up on the new sources. And I cant find any refresh/reload button in Grafana. Any advice?

[james]

@stoccafisso when changing config files inside an app, it is always a good idea to restart the app.
Did you do that?

[stoccafisso]

Yes, restarted both after updating "Status - Service discovery" in Prometheus. Even deleted the "prometheus" in Grafana and loaded a new one. It did not pick up on the new services in Prometheus

[stoccafisso]

As you see, I have two new endpoints recogniced by Prometheus, but Grafana will not pick up that information

[stoccafisso]

[james]

Can you push the metrics to the demo instance as well so I can take a look?

[james]

Also, inside Grafana, for the Prometheus Data source.
There is a pre-made dashboard for Prometheus:

Import it, and please view it so see, if you get any data at all.

[stoccafisso]

I did a brutal problemsolving hack, deleted Grafana and reinstalled it. Now it works.

I think it may have been the "Node Exporter Full" dashboard that was holding the old config in cache, I may have solved it just by deleting the Node Exporter Full and then reinstalled it.

[stoccafisso]

@james How do I do that? And is it needed? I have solved the problem?

[james]

@stoccafisso no, that is no longer needed since you have solved your issue.