Important: This documentation is about an older version. It's relevant only to the release noted, many of the features and functions have been updated or replaced. Please view the current version.

Operator and user guide

Deploy on Kubernetes

Getting started using the Helm chart

Open source

Getting started with Grafana Mimir using the Helm chart

The Helm chart allows you to configure, install, and upgrade Grafana Mimir within a Kubernetes cluster.

Before you begin

The instructions that follow are common across any flavor of Kubernetes. They also assume that you know how to install a Kubernetes cluster, and configure and operate it.

It also assumes that you have an understanding of what the kubectl command does.

Caution: This getting-started procedure is primarily aimed at local or development setups. For setting up in a production environment refer to Run Grafana Mimir in production using the Helm chart.

Hardware requirements

A single Kubernetes node with a minimum of 4 cores and 16GiB RAM

Software requirements

Kubernetes 1.20 or higher
The kubectl command for your version of Kubernetes
Helm 3 or higher

Verify that you have

Access to the Kubernetes cluster
Persistent storage is enabled in the Kubernetes cluster, which has a default storage class set up. You can change the default StorageClass.
DNS service works in the Kubernetes cluster
An ingress controller is set up in the Kubernetes cluster, for example ingress-nginx

Note: Although this is not strictly necessary, if you want to access Mimir from outside of the Kubernetes cluster, you will need an ingress. This procedure assumes you have an ingress controller set up.

Security setup

This installation will not succeed if you have enabled the PodSecurityPolicy admission controller or if you are enforcing the Restricted policy with Pod Security admission controller. The reason is that the installation includes a deployment of MinIO. The minio/minio chart is not compatible with running under a Restricted policy or the PodSecurityPolicy that the mimir-distributed chart provides.

If you are using the PodSecurityPolicy admission controller, then it is not possible to deploy the mimir-distributed chart with MinIO. Refer to Run Grafana Mimir in production using the Helm chart for instructions on setting up an external object storage and disable the built-in MinIO deployment with minio.enabled: false in the Helm values file.

If you are using the Pod Security admission controller, then MinIO and the mimir-distributed chart can successfully deploy under the baseline pod security level.

Install the Helm chart in a custom namespace

Using a custom namespace solves problems later on because you do not have to overwrite the default namespace.

Create a unique Kubernetes namespace, for example mimir-test:
console
```
kubectl create namespace mimir-test
```
For more details, see the Kubernetes documentation about Creating a new namespace.
Set up a Helm repository using the following commands:
console
```
helm repo add grafana https://grafana.github.io/helm-charts
helm repo update
```
Note: The Helm chart at https://grafana.github.io/helm-charts is a publication of the source code at grafana/mimir.
Configure an ingress:
a. Create a YAML file of Helm values called custom.yaml.
b. Add the following configuration to the file:
yaml
```
nginx:
  ingress:
    enabled: true
    ingressClassName: nginx
    hosts:
      - host: <ingress-host>
        paths:
          - path: /
            pathType: Prefix
    tls:
      # empty, disabled.
```
An ingress enables you to externally access a Kubernetes cluster. Replace <ingress-host> with a suitable hostname that DNS can resolve to the external IP address of the Kubernetes cluster. For more information, see Ingress.
Note: On Linux systems, and if it is not possible for you set up local DNS resolution, you can use the --add-host=<ingress-host>:<kubernetes-cluster-external-address> command-line flag to define the <ingress-host> local address for the docker commands in the examples that follow.
Install Grafana Mimir using the Helm chart:
bash
```
helm -n mimir-test install mimir grafana/mimir-distributed -f custom.yaml
```
Note: The output of the command contains the write and read URLs necessary for the following steps.

Check the statuses of the Mimir pods:

kubectl -n mimir-test get pods

The results look similar to this:

kubectl -n mimir-test get pods
NAME                                            READY   STATUS      RESTARTS   AGE
mimir-minio-78b59f5569-fhlhs                    1/1     Running     0          2m4s
mimir-nginx-74f8bff8dc-7kr7z                    1/1     Running     0          2m5s
mimir-distributed-make-bucket-job-z2hc8         0/1     Completed   0          2m4s
mimir-overrides-exporter-5fd94b745b-htrdr       1/1     Running     0          2m5s
mimir-query-frontend-68cbbfbfb5-pt2ng           1/1     Running     0          2m5s
mimir-ruler-56586c9774-28k7h                    1/1     Running     0          2m5s
mimir-querier-7894f6c5f9-pj9sp                  1/1     Running     0          2m5s
mimir-querier-7894f6c5f9-cwjf6                  1/1     Running     0          2m4s
mimir-alertmanager-0                            1/1     Running     0          2m4s
mimir-distributor-55745599b5-r26kr              1/1     Running     0          2m4s
mimir-compactor-0                               1/1     Running     0          2m4s
mimir-store-gateway-0                           1/1     Running     0          2m4s
mimir-ingester-1                                1/1     Running     0          2m4s
mimir-ingester-2                                1/1     Running     0          2m4s
mimir-ingester-0                                1/1     Running     0          2m4s

Wait until all of the pods have a status of Running or Completed, which might take a few minutes.

Configure Prometheus to write to Grafana Mimir

You can either configure Prometheus to write to Grafana Mimir or configure Grafana Agent to write to Mimir. Although you can configure both, you do not need to.

Make a choice based on whether or not you already have a Prometheus server set up:

For an existing Prometheus server:
1. Add the following YAML snippet to your Prometheus configuration file:
  yaml
```
remote_write:
  - url: http://<ingress-host>/api/v1/push
```
  In this case, your Prometheus server writes metrics to Grafana Mimir, based on what is defined in the existing scrape_configs configuration.
2. Restart the Prometheus server.
For a Prometheus server that does not exist yet:
1. Write the following configuration to a prometheus.yml file:
  yaml
```
remote_write:
  - url: http://<ingress-host>/api/v1/push

scrape_configs:
  - job_name: prometheus
    honor_labels: true
    static_configs:
      - targets: ["localhost:9090"]
```
  In this case, your Prometheus server writes metrics to Grafana Mimir that it scrapes from itself.
2. Start a Prometheus server by using Docker:
  bash
```
docker run -p 9090:9090  -v <absolute-path-to>/prometheus.yml:/etc/prometheus/prometheus.yml prom/prometheus
```
  Note: On Linux systems, if <ingress-host> cannot be resolved by the Prometheus server, use the additional command-line flag --add-host=<ingress-host>:<kubernetes-cluster-external-address> to set it up.

Configure Grafana Agent to write to Grafana Mimir

You can either configure Grafana Agent to write to Grafana Mimir or configure Prometheus to write to Mimir. Although you can configure both, you do not need to.

Make a choice based on whether or not you already have a Grafana Agent set up:

For an existing Grafana Agent:
1. Add the following YAML snippet to your Grafana Agent metrics configurations (metrics.configs):
  yaml
```
remote_write:
  - url: http://<ingress-host>/api/v1/push
```
  In this case, your Grafana Agent will write metrics to Grafana Mimir, based on what is defined in the existing metrics.configs.scrape_configs configuration.
2. Restart the Grafana Agent.
For a Grafana Agent that does not exist yet:
1. Write the following configuration to an agent.yaml file:
  yaml
```
metrics:
  wal_directory: /tmp/grafana-agent/wal

  configs:
    - name: agent
      scrape_configs:
        - job_name: agent
          static_configs:
            - targets: ["127.0.0.1:12345"]
      remote_write:
        - url: http://<ingress-host>/api/v1/push
```
  In this case, your Grafana Agent writes metrics to Grafana Mimir that it scrapes from itself.
2. Create an empty directory for the write ahead log (WAL) of the Grafana Agent
3. Start a Grafana Agent by using Docker:
  bash
```
docker run -v <absolute-path-to-wal-directory>:/etc/agent/data -v <absolute-path-to>/agent.yaml:/etc/agent/agent.yaml -p 12345:12345 grafana/agent
```
  Note: On Linux systems, if <ingress-host> cannot be resolved by the Grafana Agent, use the additional command-line flag --add-host=<ingress-host>:<kubernetes-cluster-external-address> to set it up.

Query metrics in Grafana

First install Grafana, and then add Mimir as a Prometheus data source.

Start Grafana by using Docker:
bash
```
docker run --rm --name=grafana -p 3000:3000 grafana/grafana
```
Note: On Linux systems, if <ingress-host> cannot be resolved by Grafana, use the additional command-line flag --add-host=<ingress-host>:<kubernetes-cluster-external-address> to set it up.
In a browser, go to the Grafana server at http://localhost:3000.
Sign in using the default username admin and password admin.
On the left-hand side, go to Configuration > Data sources.
Configure a new Prometheus data source to query the local Grafana Mimir cluster, by using the following settings:
Field Value
Name Mimir
URL http://<ingress-host>/prometheus
To add a data source, see Add a data source.
Verify success:
You should be able to query metrics in Grafana Explore, as well as create dashboard panels by using your newly configured Mimir data source. For more information, see Monitor Grafana Mimir.

Field	Value
Name	Mimir
URL	http://<ingress-host>/prometheus

Set up metamonitoring

Grafana Mimir metamonitoring collects metrics or logs, or both, about Grafana Mimir itself. In the example that follows, metamonitoring scrapes metrics about Grafana Mimir itself, and then writes those metrics to the same Grafana Mimir instance.

To enable metamonitoring in Grafana Mimir, add the following YAML snippet to your Grafana Mimir custom.yaml file:

metaMonitoring:
  serviceMonitor:
    enabled: true
  grafanaAgent:
    enabled: true
    installOperator: true
    metrics:
      additionalRemoteWriteConfigs:
        - url: "http://mimir-nginx.mimir-test.svc:80/api/v1/push"

Upgrade Grafana Mimir by using the helm command:

helm -n mimir-test upgrade mimir grafana/mimir-distributed -f custom.yaml

From Grafana Explore, verify that your metrics are being written to Grafana Mimir, by querying sum(rate(cortex_ingester_ingested_samples_total[$__rate_interval])).

Query metrics in Grafana that is running within the same Kubernetes cluster

Install Grafana in the same Kubernetes cluster.
For details, see Deploy Grafana on Kubernetes.
Stop the Grafana instance that is running in the Docker container, to allow for port-forwarding.
Port-forward Grafana to localhost, by using the kubectl command:
bash
```
kubectl port-forward service/grafana 3000:3000
```
In a browser, go to the Grafana server at http://localhost:3000.
Sign in using the default username admin and password admin.
On the left-hand side, go to Configuration > Data sources.
Configure a new Prometheus data source to query the local Grafana Mimir server, by using the following settings:
Field Value
Name Mimir
URL http://mimir-nginx.mimir-test.svc:80/prometheus
To add a data source, see Add a data source.
Verify success:
You should be able to query metrics in Grafana Explore, as well as create dashboard panels by using your newly configured Mimir data source.