container.training/slides/k8s/setup-managed.md

Deploying a managed cluster

"The easiest way to install Kubernetes is to get someone else to do it for you."
(Jérôme Petazzoni)

• Let's see a few options to install managed clusters!

• This is not an exhaustive list

  (the goal is to show the actual steps to get started)

• The list is sorted alphabetically

• All the options mentioned here require an account with a cloud provider

• ... And a credit card

---

AKS (initial setup)

• Install the Azure CLI

• Login:

  az login
  

• Select a region

• Create a "resource group":

  az group create --name my-aks-group --location westeurope
  

---

AKS (create cluster)

• Create the cluster:

  az aks create --resource-group my-aks-group --name my-aks-cluster
  

• Wait about 5-10 minutes

• Add credentials to kubeconfig:

  az aks get-credentials --resource-group my-aks-group --name my-aks-cluster
  

---

AKS (cleanup)

• Delete the cluster:

  az aks delete --resource-group my-aks-group --name my-aks-cluster
  

• Delete the resource group:

  az group delete --resource-group my-aks-group
  

• Note: delete actions can take a while too!

  (5-10 minutes as well)

---

AKS (notes)

• The cluster has useful components pre-installed, such as the metrics server

• There is also a product called AKS Engine:

  • leverages ARM (Azure Resource Manager) templates to deploy Kubernetes

  • it's "the library used by AKS"

  • fully customizable

  • think of it as "half-managed" Kubernetes option

---

Amazon EKS (the old way)

• Read the doc

• Create service roles, VPCs, and a bunch of other oddities

• Try to figure out why it doesn't work

• Start over, following an official AWS blog post

• Try to find the missing Cloud Formation template

--

.footnote[(╯°□°)╯︵ ┻━┻]

---

Amazon EKS (the new way)

• Install eksctl

• Set the usual environment variables

  (AWS_DEFAULT_REGION, AWS_ACCESS_KEY, AWS_SECRET_ACCESS_KEY)

• Create the cluster:

  eksctl create cluster
  

• Cluster can take a long time to be ready (15-20 minutes is typical)

• Add cluster add-ons

  (by default, it doesn't come with metrics-server, logging, etc.)

---

Amazon EKS (cleanup)

• Delete the cluster:

  eksctl delete cluster <clustername>
  

• If you need to find the name of the cluster:

  eksctl get clusters
  

.footnote[Note: the AWS documentation has been updated and now includes eksctl instructions.]

---

Amazon EKS (notes)

• Convenient if you have to use AWS

• Needs extra steps to be truly production-ready

• Versions tend to be outdated

• The only officially supported pod network is the Amazon VPC CNI plugin

  • integrates tightly with security groups and VPC networking

  • not suitable for high density clusters (with many small pods on big nodes)

  • other plugins should still work but will require extra work

---

Digital Ocean (initial setup)

• Install doctl

• Generate API token (in web console)

• Set up the CLI authentication:

  doctl auth init
  

  (It will ask you for the API token)

• Check the list of regions and pick one:

  doctl compute region list
  

  (If you don't specify the region later, it will use nyc1)

---

Digital Ocean (create cluster)

• Create the cluster:

  doctl kubernetes cluster create my-do-cluster [--region xxx1]
  

• Wait 5 minutes

• Update kubeconfig:

  kubectl config use-context do-xxx1-my-do-cluster
  

• The cluster comes with some components (like Cilium) but no metrics server

---

Digital Ocean (cleanup)

• List clusters (if you forgot its name):

  doctl kubernetes cluster list
  

• Delete the cluster:

  doctl kubernetes cluster delete my-do-cluster
  

---

GKE (initial setup)

• Install gcloud

• Login:

  gcloud auth init
  

• Create a "project":

  gcloud projects create my-gke-project
  gcloud config set project my-gke-project
  

• Pick a region

  (example: europe-west1, us-west1, ...)

---

GKE (create cluster)

• Create the cluster:

  gcloud container clusters create my-gke-cluster --region us-west1 --num-nodes=2
  

  (without --num-nodes you might exhaust your IP address quota!)

• The first time you try to create a cluster in a given project, you get an error

  • you need to enable the Kubernetes Engine API
  • the error message gives you a link
  • follow the link and enable the API (and billing)
    (it's just a couple of clicks and it's instantaneous)

• Clutser should be ready in a couple of minutes

---

GKE (cleanup)

• List clusters (if you forgot its name):

  gcloud container clusters list
  

• Delete the cluster:

  gcloud container clusters delete my-gke-cluster --region us-west1
  

• Delete the project (optional):

  gcloud projects delete my-gke-project
  

---

GKE (notes)

• Well-rounded product overall

  (it used to be one of the best managed Kubernetes offerings available; now that many other providers entered the game, that title is debatable)

• The cluster comes with many add-ons

• Versions lag a bit:

  • latest minor version (e.g. 1.18) tends to be unsupported

  • previous minor version (e.g. 1.17) supported through alpha channel

  • previous versions (e.g. 1.14-1.16) supported

---

Scaleway (initial setup)

• After creating your account, make sure you set a password or get an API key

  (by default, it uses email "magic links" to sign in)

• Install scw

  (you need CLI v2, which in beta as of May 2020)

• Generate the CLI configuration with scw init

  (it will prompt for your API key, or email + password)

---

Scaleway (create cluster)

• Create the cluster:

    k8s cluster create name=my-kapsule-cluster version=1.18.3 cni=cilium \
        default-pool-config.node-type=DEV1-M default-pool-config.size=3 
  

• After less than 5 minutes, cluster state will be ready

  (check cluster status with e.g. scw k8s cluster list on a wide terminal )

• Add connection information to your .kube/config file:

  scw k8s kubeconfig install `CLUSTERID`
  

  (the cluster ID is shown by scw k8s cluster list)

---

class: extra-details

Scaleway (automation)

• If you want to obtain the cluster ID programmatically, this will do it:

  scw k8s cluster list
  # or
  CLUSTERID=$(scw k8s cluster list -o json | \
        jq -r '.[] | select(.name="my-kapsule-cluster") | .id')
  

---

Scaleway (cleanup)

• Get cluster ID (e.g. with scw k8s cluster list)

• Delete the cluster:

  scw cluster delete cluster-id=$CLUSTERID
  

• Warning: as of May 2020, load balancers have to be deleted separately!

---

Scaleway (notes)

• The create command is a bit more complex than with other providers

  (you must specify the Kubernetes version, CNI plugin, and node type)

• To see available versions and CNI plugins, run scw k8s version list

• As of May 2026, Kapsule supports:

  • multiple CNI plugins, including: calico, cilium, kilo

  • Kubernetes versions 1.32 to 1.35

  • multiple container runtimes, including: Docker, containerd, CRI-O

• To see available node types and their price, check their pricing page

---

More options

• Alibaba Cloud

• IBM Cloud

• Linode Kubernetes Engine (LKE)

• OVHcloud Managed Kubernetes Service

• ...

---

Reminder...

Managed Kubernetes ≠ managed hosting!

• Running an app also involves system upgrades, supervision, on-call, backups...

• "Managed hosting" means that the hosting provider takes care of it

• In "managed Kubernetes", you are responsible for these tasks!

???

:EN:- Installing a managed cluster :FR:- Installer un cluster infogéré