Skip to content

Latest commit

 

History

History
executable file
·
137 lines (88 loc) · 3.93 KB

kops.md

File metadata and controls

executable file
·
137 lines (88 loc) · 3.93 KB
Raw

Installing a Kubernetes Cluster on Google Cloud Platform (GCP) with Kubernetes Operations (kops)

Prerequisites

Note: We will use the Google Cloud Shell which already has kubectl and gcloud tools installed.

Exercise 1: Deploy kubernetes to GCP

  1. Open the GCP console from your browser. GCP Console

  2. Login and select the assigned project or create a new one from the top menu.

  3. Enable (if not already) the Compute Engine API here GCP Console

  4. Launch the Google Cloud Shell from the top right menu.

  5. Run the following commands to set the proper region and zone:

    gcloud config set compute/region us-west1

gcloud config set compute/zone us-west1-c

    Note: Further commands will all be run in this shell.

  6. Install kops

    Run the following commands to download and install kops:

    curl -LO https://github.com/kubernetes/kops/releases/download/$(curl -s https://api.github.com/repos/kubernetes/kops/releases/latest | grep tag_name | cut -d '"' -f 4)/kops-linux-amd64

chmod +x kops-linux-amd64

mkdir -p $HOME/bin && cp kops-linux-amd64 "$_/kops"

PATH="$HOME/bin:$PATH"

echo 'PATH="$HOME/bin:$PATH"' >> ~/.bashrc

  7. Create a state store

    Run the following command and replace <unique value> with your username or other.

    BUCKET="kubernetes-<unique value>"
    gsutil mb gs://${BUCKET}

    Here we are creating a GCP bucket to store the cluster configuration for kops

  8. Create the cluster configuration with calico.

    Run the following commands.

    PROJECT=`gcloud config get-value project`

export KOPS_FEATURE_FLAGS=AlphaAllowGCE # to unlock the GCE features

kops create cluster simple.k8s.local --zones us-west1-c --state gs://${BUCKET} --project=${PROJECT}

    Note: This only created the configuration which can be viewed here:

    kops get cluster --state gs://${BUCKET}

kops get instancegroup --state gs://${BUCKET} --name simple.k8s.local

kops get cluster --state gs://${BUCKET} simple.k8s.local -oyaml

    Note: We can set a variable for the state store to make commands shorter:

    export KOPS_STATE_STORE=gs://${BUCKET}

echo "export BUCKET=${BUCKET}" >> ~/.bashrc
echo 'export KOPS_STATE_STORE=gs://${BUCKET}' >> ~/.bashrc
echo 'export KOPS_FEATURE_FLAGS=AlphaAllowGCE' >> ~/.bashrc

  9. Building the cluster in GCE

    Run the following command and confirm the output

    kops update cluster simple.k8s.local

    To proceed with the operation we need to confirm the command by running

    kops update cluster simple.k8s.local --yes

    Validate

    kops validate cluster

    After a few minutes the cluster will be ready and can be viewed from kubectl

    kubectl cluster-info

kubectl get nodes --show-labels
    gcloud compute ssh $(gcloud compute instances list|awk '/master/ {print $1}')

Exercise 2 (Optional): Identify resources that have been created

  1. In GCE Cloud Console find and investigate the following resources, that have been created as a result of previous exercises
    • Compute Engine -> VM instances
    • Compute Engine -> Disks
    • Compute Engine -> Instance groups
    • Compute Engine -> Instance templates
    • VPC Network -> External IP addresses
    • VPC Network -> Firewall rules
    • Network services -> Load balancing We will examine all created infrastructure in details on day 3.

Next: Pods