Tutorial: Getting started
The tutorial consists of the following steps:
Prerequisites
Charmed Openstack and COS
To get started, you will need a Charmed OpenStack deployment with COS.
Openstack-exporter snap (revisions ≤ 27)
Charm revisions 27 or older require a custom built snap from the branch at openstack-exporter/openstack-exporter#355, uploaded to the charm via a resource. A temporary pre-built snap can be found in the assets section at Release Revision 2 · canonical/openstack-exporter-operator · GitHub .
wget https://github.com/canonical/openstack-exporter-operator/releases/download/rev2/golang-openstack-exporter_amd64.snap
Deploy the charm
Switch to the Juju openstack model. Then you can deploy the charm:
# Revisions 27 or older
juju deploy openstack-exporter --resource openstack-exporter=golang-openstack-exporter_amd64.snap
# Revisions > 27
juju deploy openstack-exporter
The charm must be integrated with keystone (for credentials) and grafana-agent (to send Grafana dashboards to Grafana and metrics to Prometheus).
juju integrate openstack-exporter:credentials keystone:identity-admin
juju integrate openstack-exporter:cos-agent grafana-agent:cos-agent
If your OpenStack cloud is configured with SSL, with a local CA, you may need to set the ssl_ca option in the charm.
For revisions 27 or older, the cache mechanism for the metrics are disabled by default which can cause time out issues and metrics not propagated into COS if the OpenStack cloud is a large deployment.
On revisions > 27 the cache mechanism for the metrics are enabled by default.
The charm should settle to active/idle status:
$ juju status openstack-exporter
...
App Version Status Scale Charm Channel Rev Exposed Message
grafana-agent active 1 grafana-agent latest/edge 96 no
openstack-exporter active 1 openstack-exporter latest/edge 10 no
Unit Workload Agent Machine Public address Ports Message
openstack-exporter/0* active idle 15 10.5.0.233
grafana-agent/0* active idle 10.5.0.233
View the Grafana dashboards
Navigate to the Grafana web ui and click the dashboards tab. Under the “General” folder, you should see the three dashboards:
See the Grafana Dashboards section of the docs for more information about the provided dashboards.
Example screenshot from the OpenStack Hypervisor Overview dashboard, displaying information about cpu, memory, and disk usages for the hypervisors, and stats on instances running on the cloud:
