Skip to main content

How to vertically scale Redis

This guide explains how to adjust the CPU, memory, and storage resources of your Redis instance on Hikube, either via a predefined preset or by defining explicit values.

Prerequisites​

  • A Redis instance deployed on Hikube (see the quick start)
  • kubectl configured to interact with the Hikube API
  • The YAML configuration file for your Redis instance

Steps​

1. Check current resources​

Review the current configuration of your Redis instance:

kubectl get redis my-redis -o yaml

Note the values of resourcesPreset, resources, replicas and size in the spec section.

2. Option A: Modify the resourcesPreset​

The simplest way to scale is to use a predefined preset. Here are the available presets:

PresetCPUMemory
nano250m128Mi
micro500m256Mi
small1512Mi
medium11Gi
large22Gi
xlarge44Gi
2xlarge88Gi

For example, to go from nano to medium:

redis-medium.yaml
apiVersion: apps.cozystack.io/v1alpha1
kind: Redis
metadata:
  name: my-redis
spec:
  replicas: 2
  resourcesPreset: medium
  size: 2Gi
  authEnabled: true

3. Option B: Define explicit resources​

For precise control, specify CPU and memory directly with the resources field:

redis-custom-resources.yaml
apiVersion: apps.cozystack.io/v1alpha1
kind: Redis
metadata:
  name: my-redis
spec:
  replicas: 2
  resources:
    cpu: 2000m
    memory: 4Gi
  size: 5Gi
  authEnabled: true
warning

If the resources field is defined, the resourcesPreset value is entirely ignored. Remove resourcesPreset from the manifest to avoid confusion.

4. Adjust the number of replicas if needed​

You can also increase the number of replicas to distribute the read load:

redis-scaled.yaml
apiVersion: apps.cozystack.io/v1alpha1
kind: Redis
metadata:
  name: my-redis
spec:
  replicas: 3
  resourcesPreset: large
  size: 5Gi
  storageClass: replicated
  authEnabled: true

5. Apply the update​

kubectl apply -f redis-medium.yaml
tip

Redis is an in-memory data store: the allocated memory (resources.memory or that of the preset) must be sufficient to hold your entire dataset. Monitor memory usage before scaling.

Verification​

Verify that the resources have been updated:

# Verifier la configuration de la ressource Redis
kubectl get redis my-redis -o yaml | grep -A 5 resources

# Verifier l'etat des pods Redis
kubectl get pods -l app.kubernetes.io/instance=my-redis

Expected output:

NAME              READY   STATUS    RESTARTS   AGE
my-redis-0        1/1     Running   0          2m
my-redis-1        1/1     Running   0          2m

Going further​