Skip to content

Latest commit

 

History

History
305 lines (195 loc) · 9.18 KB

lab1b.md

File metadata and controls

305 lines (195 loc) · 9.18 KB
Raw

Lab 1 - Set-up OpenFaaS with Kubernetes

Install latest kubectl

Install kubectl for your operating system using the instructions below or the official documentation

  • Linux
export VER=$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)
curl -LO https://storage.googleapis.com/kubernetes-release/release/$VER/bin/linux/amd64/kubectl
chmod +x kubectl
mv kubectl /usr/local/bin/
  • MacOS
export VER=$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)
curl -LO https://storage.googleapis.com/kubernetes-release/release/$VER/bin/darwin/amd64/kubectl.exe
chmod +x kubectl
mv kubectl /usr/local/bin/
  • Windows
export VER=$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)
curl -LO https://storage.googleapis.com/kubernetes-release/release/$VER/bin/windows/amd64/kubectl.exe
chmod +x kubectl.exe
mkdir -p $HOME/bin/
mv kubectl $HOME/bin/

Setup a Kubernetes cluster

You can follow the labs whilst using Kubernetes, but you may need to make some small changes along the way. The service address for the gateway changes from http://gateway:8080 to http://gateway.openfaas:8080. As far as possible these differences have been documented and alternatives are provided in each lab.

Create a local cluster on your laptop

k3s using k3d

If you have Docker on your computer, then you can use k3d from Rancher Labs. It installs a lightweight version of Kubernetes called k3s and runs it within a Docker container, meaning it will work on any computer that has Docker.

$ k3d create
INFO[0000] Created cluster network with ID 9a7d5887754d3e317b5c1500f706a5ae602077a18bc71bcedb9fae86ebd84c0b 
INFO[0000] Created docker volume  k3d-k3s-default-images 
INFO[0000] Creating cluster [k3s-default]               
INFO[0000] Creating server using docker.io/rancher/k3s:v0.9.1... 
INFO[0000] SUCCESS: created cluster [k3s-default]

Switch into the k3d context:

export KUBECONFIG="$(k3d get-kubeconfig --name='k3s-default')"
kubectl cluster-info

Note: You have to run this on any new terminal you open.

Docker for Mac

Note that Kubernetes is only available in Docker for Mac 17.12 CE and higher

With Minikube

  • To install Minikube download the proper installer from latest release depending on your platform.

  • Now run Minikube with

$ minikube start

The minikube VM is exposed to the host system via a host-only IP address. Check this IP with minikube ip. This is the IP you will later use for the gateway URL.

Note: Minikube also requires a Hypervisor such as VirtualBox or Hyperkit (on MacOS). Follow the minikube instructions and documentation

Create a remote cluster on the cloud

You can create a remote cluster in the cloud and enjoy the same experience as if you were developing locally whilst saving on RAM/CPU and battery. The costs for running a cluster for 1-2 days is minimal.

Run on DigitalOcean's Kubernetes Service

You can use free credits to create a cluster through DigitalOcean's UI.

The DigitalOcean dashboard will then guide you through how to configure your kubectl and KUBECONFIG file for use in the labs.

Even if you have already claimed free credit, the running costs for a 2-3 node cluster for 24-48 hours is negligible.

  • Click on Kubernetes on the left panel of the dashboard and then click "Enable Limited Access"

  • Once logged in, click the Kubernetes menu item and create a Cluster.

It is recommended to use the latest Kubernetes version available and the to select your nearest Datacenter region to minimize latency.

  • Under "Add node pool(s)"

Use 2x 4GB / 2vCPU

Note: You can add more capacity at a later time, if required

Keep track of your API key (copy it to clipboard)

  • Authenticate the CLI
$ doctl auth init

Paste in your API key

  • Now get the cluster's name:
$ doctl k8s cluster list
GUID    workshop-lon1      nyc1      1.13.5-do.1    provisioning    workshop-lon1-1
  • Save a config file so that kubectl is pointing at the new cluster:
$ doctl k8s cluster kubeconfig save workshop-lon1

You now need to switch your Kubernetes context to point at the new cluster.

Find the cluster name with kubectl config get-contexts, if it's not highlighted type in kubectl config set-context <context-name>.

Run on GKE (Google Kubernetes Engine)

Login into Google Cloud, create a project and enable billing for it. If you don’t have an account you can sign up here for free credits.

Install the Google Cloud SDK - this will make the gcloud and kubectl commands available. For Windows follow the instructions from the documentation.

After installing the gcloud command line utility, configure your project with gcloud init and set the default project, compute region and zone (replace PROJECT_ID with your own project):

$ gcloud config set project PROJECT_ID 
$ gcloud config set compute/region us-central1
$ gcloud config set compute/zone us-central1-a

Enable the Kubernetes service:

$ gcloud services enable container.googleapis.com

Install kubectl:

gcloud components install kubectl

Create a Kubernetes cluster:

$ gcloud container clusters create openfaas \
--zone=us-central1-a \
--num-nodes=1 \
--machine-type=n1-standard-2 \
--disk-size=30 \
--no-enable-cloud-logging

Set up credentials for kubectl:

$ gcloud container clusters get-credentials openfaas

Create a cluster admin role binding:

$ kubectl create clusterrolebinding "cluster-admin-$(whoami)" \
--clusterrole=cluster-admin \
--user="$(gcloud config get-value core/account)"

Now verify kubectl is configured to the GKE cluster:

$ kubectl get nodes
NAME                                   STATUS    ROLES     AGE       VERSION
gke-name-default-pool-eceef152-qjmt   Ready     <none>    1h        v1.10.7-gke.2

Deploy OpenFaaS

The instructions for deploying OpenFaaS change from time to time as we strive to make this even easier.

Install OpenFaaS

There is a new tool called k3sup which can install helm charts, including OpenFaaS. This makes the installation procedure much faster.

Install with k3sup

  • Get k3sup

For MacOS / Linux:

curl -SLsf https://get.k3sup.dev/ | sudo sh

For Windows:

curl -SLsf https://get.k3sup.dev/ | sh
  • Install the OpenFaaS app

If you're using a managed cloud Kubernetes service which supplies LoadBalancers, then run the following:

k3sup app install openfaas --load-balancer

Note: the --load-balancer flag has a default of false, so by passing the flag, the installation will request one from your cloud provider.

If you're using a local Kubernetes cluster or a VM, then run:

k3sup app install openfaas

Or install with helm (advanced)

If you prefer, you can install OpenFaaS using the helm chart instructions.

Log into your OpenFaaS gateway

  • Check the gateway is ready
kubectl rollout status -n openfaas deploy/gateway

If you're using your laptop, a VM, or any other kind of Kubernetes distribution run the following instead:

kubectl port-forward svc/gateway -n openfaas 8080:8080

This command will open a tunnel from your Kubernetes cluster to your local computer so that you can access the OpenFaaS gateway. There are other ways to access OpenFaaS, but that is beyond the scope of this workshop.

Your gateway URL is: http://127.0.0.1:8080

If you're using a managed cloud Kubernetes service then get the LoadBalancer's IP address or DNS entry from the EXTERNAL-IP field from the command below.

kubectl get svc -o wide gateway-external -n openfaas

Your URL will be the IP or DNS entry above on port 8080.

  • Log in:
export OPENFAAS_URL="" # Populate as above

# This command retrieves your password
PASSWORD=$(kubectl get secret -n openfaas basic-auth -o jsonpath="{.data.basic-auth-password}" | base64 --decode; echo)

# This command logs in and saves a file to ~/.openfaas/config.yml
echo -n $PASSWORD | faas-cli login --username admin --password-stdin
  • Check that faas-cli list works:
faas-cli list

Permanently save your OpenFaaS URL

Edit ~/.bashrc or ~/.bash_profile - create the file if it doesn't exist.

Now add the following - changing the URL as per the one you saw above.

export OPENFAAS_URL="" # populate as above

Now move onto Lab 2