Install Grafana Loki⚓︎

EDP configures the logging with the help of Grafana Loki aggregation system.

Installation⚓︎

To install Loki, follow the steps below:

Create logging namespace:
```
  kubectl create namespace logging
```
Note

On the OpenShift cluster, run the oc command instead of the kubectl command.
Add a chart repository:
```
  helm repo add grafana https://grafana.github.io/helm-charts
  helm repo update
```
Note

It is possible to use Amazon Simple Storage Service Amazon S3 as an object storage for Loki. To configure access, please refer to the IRSA for Loki documentation.

Install Loki v.2.6.0:

  helm install loki grafana/loki \
  --version 2.6.0 \
  --values values.yaml \
  --namespace logging

Check out the values.yaml file sample of the Loki customization:

View: values.yaml

image:
  repository: grafana/loki
  tag: 2.3.0
config:
  auth_enabled: false
  schema_config:
    configs:
    - from: 2021-06-01
      store: boltdb-shipper
      object_store: s3
      schema: v11
      index:
        prefix: loki_index_
        period: 24h
  storage_config:
    aws:
      s3: s3://<AWS_REGION>/loki-<CLUSTER_NAME>
    boltdb_shipper:
      active_index_directory: /data/loki/index
      cache_location: /data/loki/boltdb-cache
      shared_store: s3
  chunk_store_config:
    max_look_back_period: 24h
resources:
  limits:
    memory: "128Mi"
  requests:
    cpu: "50m"
    memory: "128Mi"
serviceAccount:
  create: true
  name: edp-loki
  annotations:
    eks.amazonaws.com/role-arn: "arn:aws:iam::<AWS_ACCOUNT_ID>:role/AWSIRSA‹CLUSTER_NAME›‹LOKI_NAMESPACE›Loki
persistence:
  enabled: false

Note

In case of using cluster scheduling and amazon-eks-pod-identity-webhook, it is necessary to restart the Loki pod after the cluster is up and running. Please refer to the Schedule Pods Restart documentation.

Configure custom bucket policy to delete the old data.