Redis-stack helm chart

Installation

To add the repo:

helm repo add redis-stack https://redis-stack.github.io/helm-redis-stack/

To install redis-stack helm chart with latest images, run:

helm install redis-stack redis-stack/redis-stack

To install redis-stack-server helm chart with latest images, run:

helm install redis-stack redis-stack/redis-stack-server

To install the helm chart with specific redis tag, just add --set tag:

helm install redis-stack redis-stack/redis-stack --set redis_stack.tag="<TAG>"

For example, to run redis stack with redis version 7.0.0, run:

helm install redis-stack redis-stack/redis-stack --set redis_stack.tag="7.0.0-RC5"

Usage

To connect to redis-cli run:

kubectl get pods -o name | grep redis-stack

Copy the name of the pod (E.g. redis-stack-77d596476d-6j4d2) and run the following command:

kubectl exec -it <POD_NAME> -- redis-cli

By default redis-stack will have no password, Redis Stack supports the ability to configure multiple named users, each with their own password and access control configuration. Refer to the Redis Access Control List documentation for more information. Alternatively the configuration file can be modified and the requirepass directive added. This can also be triggered via the REDIS_ARGS environment variable. Examples are available on the docker image page

Production resource tuning

When deploying the production-ready redis-stack-server chart, pinning CPU/memory requests and limits is critical so your scheduler can reserve the right capacity. The chart exposes the redis_stack_server.resources block in charts/redis-stack-server/values.yaml:

redis_stack_server:
  resources:
    requests:
      cpu: 100m
      memory: 1Gi
    limits:
      cpu: 500m
      memory: 2Gi

Adjust these values for your node sizes (e.g. cpu: 250m requests and cpu: 1 limit with proportional memory) either by editing the values file or overriding via --set redis_stack_server.resources.requests.memory=.... Apply the tuned profile with something like helm upgrade --install redis-stack redis-stack/redis-stack-server -f values-production.yaml.

Verify the pod receives the requests/limits you expect using kubectl describe pod <name> before pushing to production.