CVMFS/Frontier squid on Kubernetes

Build the image

docker build -t squid .

then push to a repository and adjust the image name in squid-deployment.yaml as appropriate. We use a deployment to specify that a certain number of squid instances should be running:

kubectl create -f squid-deployment.yaml

Create the squid service:

kubectl create -f squid-service.yaml

which enables the squid(s) to be accessed from within the Kubernetes cluster at http://squid:3128.

A quick check to see that everything is working as expected, in particular that the appropriate number of squid pods are running:

$ kubectl get deployments,pods,services
NAME              DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
deploy/squid      1         1         1            1           26s
NAME                                READY     STATUS    RESTARTS   AGE
po/squid-725328492-q7qzk            1/1       Running   0          26s
NAME                           CLUSTER-IP      EXTERNAL-IP   PORT(S)    AGE
svc/kubernetes                 10.75.240.1     <none>        443/TCP    7d
svc/squid                      10.75.241.169   <none>        3128/TCP   17h