Skip to content

Latest commit

 

History

History
62 lines (53 loc) · 1.72 KB

digital_ocean.md

File metadata and controls

62 lines (53 loc) · 1.72 KB

Alternative binderhub-build on Digital Ocean

Instead of azure-cli

we need doctl (Digital Ocean ConTroL)

wget https://github.com/digitalocean/doctl/releases/download/v1.13.0/doctl-1.13.0-linux-amd64.tar.gz
tar -xvf doctl-1.13.0-linux-amd64.tar.gz
sudo mv doctl /usr/bin/

Instead of az-login

You'll need an authentication token. Login to digital ocean online. Select API and then generate a new token. Give it a cool name, like 'bindertime'.
doctl auth init

Create your cluster!

digital ocean's k8s support from the command line is in beta, but it works fine export DIGITALOCEAN_ENABLE_BETA=1`

doctl k8s cluster create bindertime-k8s --region lon1 --version 1.12.1-do.2 --node-pool="name=worker-pool;count=2"

Save config file

Make a directory to store your config
mkdir binderhub; cd binderhub
doctl k8s cluster kubeconfig show bindertime-k8s > config.yaml

Check we are up and running

kubectl --kubeconfig='config.yaml' get node

Setting up Helm

Same as in azure case, but you might need the --kubeconfig='config.yaml' flag.

Automate creation of the config files

docker_id=alexmorley
docker_pass=*********
docker_prefix=alex-binder-

token1=`openssl rand -hex 32`
token2=`openssl rand -hex 32`
cat <<EOF > secret.yaml
jupyterhub:
  hub:
    services:
      binder:
        apiToken: $token1
  proxy:
    secretToken: $token2
registry:
  username: $docker_id
  password: $docker_pass
EOF

and then:

cat <<EOF > config.yaml
config:
  BinderHub:
    use_registry: true
    image_prefix: $docker_id/$docker_prefix
EOF