Ctrlk

Grafana Monitoring Installation and Configuration on RHOSO18

1] Clone triliovault-cfg-scripts Repository

The following steps must be performed on the bastion node of an already installed RHOSO environment.

Run the following command to clone the triliovault-cfg-scripts GitHub repository:

git clone -b 6.1.7-1 https://github.com/trilioData/triliovault-cfg-scripts.git
cd triliovault-cfg-scripts/redhat-director-scripts/monitoring-openshift/

2] Enable User Workload Monitoring in OpenShift

2.1] Check if user workload monitoring is already enabled on OpenShift

  • Check if the following ConfigMap exists:

oc get cm -n openshift-monitoring | grep cluster-monitoring-config

  • If it exists, check whether the enableUserWorkload: true parameter is present with its value set to true under config.yaml, as shown below. If it is present, user workload monitoring is enabled. If it is not present or is set to false, you must enable it using the next steps. Otherwise, skip the steps to enable user workload monitoring.

oc get -n openshift-monitoring cm cluster-monitoring-config -o yaml
apiVersion: v1
data:
  config.yaml: |
    enableUserWorkload: true
kind: ConfigMap
metadata:
  name: cluster-monitoring-config
  namespace: openshift-monitoring

2.2] If user workload monitoring is not enabled, enable it

  • Check whether the following ConfigMap already exists. If it does not exist, create it using the following YAML file. No changes are required in the YAML file.

  • If the ConfigMap cluster-monitoring-config already exists, add the following parameter to it:

  • Edit the ConfigMap using:

3] Steps to Install Prometheus SQL Exporter

3.1] Create SQL Exporter ConfigMap

  • First, you need to create a ConfigMap that holds the SQL Exporter's configuration. This includes the database connection string.

  • Edit sql-exporter-config.yaml. Before applying, locate the data_source_name field and update it with the correct MySQL username, password, and hostname for your database.

  • Important: Do not change the database name or the overall format of the data_source_name string. Only update credentials and hostname.

  • Create the Secret.

Before applying the deployment, create the secret containing the DSN string:

Replace <username>, <password>, and <hostname> with the actual values.

  • Apply the ConfigMap:

3.2] Apply SQL Exporter Deployment

  • Deploy the SQL Exporter application using its Deployment manifest:

  • Verify pod status after deployment:

You should see a pod with the status Running.

3.3] Apply SQL Exporter Service

  • Create a Kubernetes Service to expose the SQL Exporter within the cluster so that other services (such as Prometheus) can discover and scrape its metrics:

  • Verify Service Status: Confirm that the Service has been created.

3.4] Apply SQL Exporter ServiceMonitor

Finally, create a ServiceMonitor resource. This tells Prometheus (if configured to discover ServiceMonitors in this namespace) to automatically scrape metrics from your SQL Exporter Service.

4] Steps to Install Prometheus OpenStack Exporter

Follow below steps to deploy the OpenStack Exporter:

4.1] Create ConfigMap with OpenStack Cloud Credentials

First, create a ConfigMap containing your OpenStack cloud credentials. This ConfigMap will be mounted into the exporter pod.

Replace all placeholder variables (such as auth URL, username, password, etc.) with actual values from your OpenStack cloud in the following config file.

4.2] Create Keystone CA Certificate Secret

Before deploying the OpenStack Exporter, create the required secret containing the Keystone CA certificate.

4.3] Deploy OpenStack Exporter

Next, deploy the OpenStack Exporter application using its Deployment manifest.

4.4] Verify OpenStack Exporter Pod Status

After deployment, verify that the exporter pod is running correctly.

You should see a pod with a status of Running.

4.5] Check OpenStack Exporter Pod Logs

To ensure the exporter is starting and connecting to OpenStack successfully, check its logs. Replace <openstack-exporter-pod-name> with the actual name of your exporter pod obtained from the previous step.

Look for messages indicating a successful connection to OpenStack and that it is listening for metrics requests (for example, on port 9180).

4.6] Create Service for OpenStack Exporter

Create a Kubernetes Service to expose the OpenStack Exporter within the cluster, allowing other services (like Prometheus) to discover and scrape its metrics.

4.7] Create ServiceMonitor for Prometheus Integration

Finally, create a ServiceMonitor resource. This tells Prometheus (if configured to discover ServiceMonitors in this namespace) to automatically scrape metrics from your OpenStack Exporter Service.

5] Install and Configure Grafana

5.1] Install Grafana Operator

  1. Go to Administrator View → Operators → OperatorHub.

  2. Search for Grafana Operator and select it.

  3. Click Install to open the operator installation page.

  4. In the installation configuration:

    • Choose the appropriate Update Channel.

    • Select the latest available operator version.

    • Under Installation Mode, select:

      • A specific namespace on the cluster

    • Select Installed Namespace to:

    • Keep Update Approval as Automatic (recommended).

  5. Click Install to confirm the installation.

Note: It is recommended to install the Grafana Operator in the openshift-user-workload-monitoring namespace, as this namespace is used for user workload monitoring in RHOSO.

5.2] Create Grafana service account

5.3] Create credential secret for Grafana

Edit the following secret file and set fresh username and password values.

Token must be fetched using the following command:

After running the above command, the token is copied into the Grafana secret YAML file.

5.4] Create Grafana instance

For a fresh deployment, create a Grafana instance:

In case of a token update on an existing Grafana instance, this is not required if you are installing Grafana for the first time:

5.5] Verify Grafana installation

5.6] Expose Grafana service if route not found

If the above command to get the Grafana route returns a NotFound error, you need to expose the Grafana service manually to create the route:

5.7] Create Grafana datasource

Create Grafana datasource to fetch metrics from the cluster monitoring namespace

Verify

PreviousGrafana MonitoringNextProcedure to Import a Grafana Dashboard

Last updated

Was this helpful?