#Adding Grafana plugins (or another solution)

1 messages · Page 1 of 1 (latest)

shut fulcrum
#

Hi there, I have a customer who requests a weekly status (per email) with performance stats. I have created the dashboard, but the only way Grafana can report is via a URL... and this customer does not have access to the grafana view. I tried to install "Grafana Image Renderer" from the cmd-line (docker -exec ....) but that broke grafana in a way, so that it was unable to restart... so I reverted back from a snapshot.
Is there others who has the same requirements and have solved it differently? Basically an URL isn't working here, we need either jpeg or pdf and some kind of report generator would be great.
I was thinking that since grafana broke by installing the plugin, maybe it would be possible to install grafana on another machine (without docker involved), and connect to the NAbox?

shell nebula
#

You can add NABox Prometheus to a separate Grafana instance outside of NABox.

URL: https://nabox_ip/prometheus
Add basic auth
Skip TLS verify to false if needed.

silk edge
#

We can also try to understand what is happening with the plugin. Let me do some tests.

shut fulcrum
silk edge
#

Understood, I'll do some test see what happened

#

Ok Grafana needs permissions on /usr/local/nabox/files/grafana/provisioning/plugins/ that's why it is crashing after install

#
mkdir /usr/local/nabox/files/grafana/provisioning/plugins/
chown 472 /usr/local/nabox/files/grafana/provisioning/plugins/
dc exec grafana grafana-cli plugins install grafana-image-renderer
#

With that said...
Grafana Docker image

This plugin is not compatible with the current Grafana Docker image and requires additional system-level dependencies. We recommend setting up another Docker container for rendering and using remote rendering instead. For instruction, refer to Run in Docker.

If you still want to install the plugin with the Grafana Docker image, refer to the instructions on building a custom Grafana image in Grafana Docker documentation.

#

So I think a better option is to configure docker-compose.override.yaml to add this service as a separate container, like described in "Run in Docker"