Install Kubectl
sudo apt update && sudo apt upgrade -y
sudo apt-get install -y ca-certificates curl
sudo apt-get install -y apt-transport-https
sudo curl -fsSLo /etc/apt/keyrings/kubernetes-archive-keyring.gpg https://packages.cloud.google.com/apt/doc/apt-key.gpg
echo "deb [signed-by=/etc/apt/keyrings/kubernetes-archive-keyring.gpg] https://apt.kubernetes.io/ kubernetes-xenial main" | sudo tee /etc/apt/sources.list.d/kubernetes.list
sudo apt-get update
sudo apt-get install -y kubectl
Install Helm
sudo snap install helm --classic
Note: Make sure you're connected to cluster
Helm install <app-name> Chart/
Application should now be deployed.
Each Redis Replica runs on it’s own node, connected to the wp-redis-<main>-0 instance.
The different kubernetes services will assign an internal cluster-ip to the wp-redis-<main> service. It is this service we want to connect our WordPress service to. Once the WordPress is connected to the Redis service, the object cache will immediately optimize application performance.
Bootstrap Redis Cluster
Using the bitnami/redis chart:
helm install <my-redis-release> \
--set auth.password=secretpassword \
oci://registry-1.docker.io/bitnamicharts/redis
NOTE:
helm install <my-redis-release> uses the helm (kubernetes) package manager to set the name of the installation.
The command --set auth.password=<secret-password> is an override that sets a password to access the database.
The command oci://registry-1.docker.io/bitnamicharts/redis is pulling from a specific redis repository. This can be any registry, and doesn’t necessarily have to be the one I used. YMMV.
There will be some output instructions after running this install. It is important to execute the commands.
The first command exports the secretpassword as an environmental variable.
export REDIS_PASSWORD=$(kubectl get secret --namespace default wp-redis -o jsonpath="{.data.redis-password}" | base64 -d)
The next steps are what enable your cluster to be reachable.
kubectl run --namespace default redis-client --restart='Never' --env REDIS_PASSWORD=$REDIS_PASSWORD --image
--command -- sleep infinity
kubectl exec --tty -i redis-client \ --namespace default -- bash
sudo apt install redis-tools
REDISCLI_AUTH="$REDIS_PASSWORD" redis-cli -h <my-redis-release>-master
REDISCLI_AUTH="$REDIS_PASSWORD" redis-cli -h <my-redis-release>-replicas
kubectl port-forward --namespace default svc/<my-redis-release>-master 6379:6379 &
REDISCLI_AUTH="$REDIS_PASSWORD" redis-cli -h 127.0.0.1 -p 6379
Your Redis Cluster is now ready to receive connections. Next, we will manually update the WordPress pod to connect it to the Redis Cluster.
kubectl get pods
kubectl exec -it <wordpress-pod> -- bash
apt-get update && apt-get upgrade -y \
apt-get install vim
vim wp-config.php
define('WP_REDIS_PASSWORD','<your-password>');
define('WP_REDIS_HOST','<my-redis-release>-master');
define('WP_REDIS_PORT','6379');
Save and exit pod.
From the WordPress-Site/wp-admin page, install Redis Object Cache. Once installed, you should be able to enable object cache and get a successmessage.
I have included some simple scripts to benchmark the application to test with cache enabled compared with site performance with cahce not enabled. This should run with approximately ~250 rps comfortably on a single node. You'll need to install the k6 binary.
run k6 run <benchmark-name>.js
helm repo add datadog https://helm.datadoghq.com
helm install wp-datadog -f datadog-values.yaml --set datadog.apiKey='<api-key>' datadog/datadog
helm upgrade -f datadog-values.yaml wp-datadog datadog/datadog
helm upgrade -f datadog-values.yaml wp-datadog datadog/datadog
datadog-agent status
This will allow you to have observability of Kubernetes cluster.