This is archived documentation for v1.4.2. Go to the latest version.

Set up a GEM instance

Create a Grafana Enterprise Metrics instance:

Create an instance (or in Cortex terms, a tenant):

  1. Navigate to Grafana Metrics Enterprise > Instances.

  2. Click on Create new instance.

  3. Choose a name for this very instance. For demonstration purposes, use the name dev-instance.

  4. Select the cluster.

  5. Click Create instance.


Create an access policy

Create a data source access policy and token, which are used by Grafana Enterprise itself to access the metrics in the instance named dev-instance:

  1. Navigate to Grafana Enterprise Metrics > Access policies.

  2. Click Create new access policy.

  3. Choose a name for the policy. For demonstration purposes, use dev-read-write-policy.

  4. Enable the scope metrics:read and metrics:write.

  5. Select the instance dev-instance.

  6. Click on Create. create-ap

  7. From the newly created access policy, click + Add token.

  8. Name the token dev-token, and click on Create.

  9. In the next window, copy the token.

You can add the following example of a remote-write configuration to your Prometheus or Grafana Cloud Agent configuration files:

  - url: http://metrics-enterprise/api/v1/push
      username: dev-instance
      password: ZGV2LXJlYWQtd3JpdGVyLXBvbGljeS1kZW1vLXRva2VuOjY/ezduMTVhJDQvPGMvLzQ1SzgsJjFbMQ==

Setup remote-write to your instance

To enable writes to your cluster, add the above remote-write configuration snippet to the configuration file of an existing Prometheus or Grafana Cloud Agent. If you do not have an existing metrics collector you can get started with the Grafana Cloud Agent.

An example agent configuration would be:

  log_level: info
  http_listen_port: 8081

  wal_directory: /tmp/agent
    scrape_interval: 5s
    - name: default
        - job_name: metrics-enterprise
            - targets: ["localhost:8080"]
        - job_name: agent
            - targets: ["localhost:8081"]

    enabled: true
    enabled: true
    - url: http://<gme-hostname>/api/v1/push
        username: dev-instance
        password: ZGV2LXJlYWQtd3JpdGVyLXBvbGljeS1kZW1vLXRva2VuOjY/ezduMTVhJDQvPGMvLzQ1SzgsJjFbMQ==

Setup your instance as a Grafana datasource

Following that we can use this access policy and token to create a new datasource in Grafana:

  1. Navigate to Configuration ≫ Data Sources.

  2. Click on Add new data source.

  3. Select Prometheus.

  4. Set the URL to http://<gme-host>/prometheus.

  5. Enable Basic Auth and use User dev-instance and as Password the token from your clipboard.

  6. Click Save & Test.


Visualize your data

Once you have created a datasource you should now be able to visualize your metrics in the Grafana Explore page.