Monitoring: Grafana

If you added metrics collection support to your Anbox Cloud installation you can access the Grafana service to get your first insight into the active cluster.

Grafana is a tool for analytics and monitoring that allows to query, visualize or alert based on the metrics of the cluster or individual containers.

Every cluster LXD node contains a running Telegraf instance gathering machine metrics.
Besides, Anbox containers existing in that node also report their metrics to that same Telegraf instance.

Grafana takes its data by considering the Telegraf instances as metrics targets and polls for new content

metrics

Access Grafana

Open your favourite browser and type:

https://[Apache_IP]/grafana

where Apache_IP is the IP of the machine the grafana-proxy application is deployed to.

You can check the Apache IP in a juju status output executed from command line:

$ juju status

Once entered the above URL, You are now redirected to the login page requiring to enter user and password. In order to obtain the administrator credential open a console and type:

$ juju run-action --wait grafana/0 get-admin-password

Note: You have to replace the 0 in grafana/0 with the unit version of the installed grafana juju unit.

You should obtain a response like:

unit-grafana-0:
  id: 29f07367-556b-41d0-8318-b9fa13a78b63
  results:
    password: jd673zyYWkR7kyPW
  status: completed
  timing:
    completed: 2018-06-29 13:29:38 +0000 UTC
    enqueued: 2018-06-29 13:29:36 +0000 UTC
    started: 2018-06-29 13:29:38 +0000 UTC
  unit: grafana/0

From that response we get the password to login. In the browser on the Grafana website you can now type in the credentials:

User: admin
Password: jd673zyYWkR7kyPW

Note: The value jd673zyYWkR7kyPW is illustrative. Replace it by the one received as response
to the juju run-action command