Dec
08
2019
Kubernetes as a Service with Rancer2 at Hetzner using Terraform and Helm

In this blog post you will find my personal notes on how to setup a Kubernetes as a Service (KaaS). I will be using Terraform to create the infrastructure on Hetzner’s VMs, Rancher for KaaS and Helm to install the first application on Kubernetes.

rke_k8s.png

Many thanks to dear friend: adamo for his help.

Terraform

Let’s build our infrastructure!
We are going to use terraform to build 5 VMs

  • One (1) master
  • One (1) etcd
  • Two (2) workers
  • One (1) for the Web dashboard

I will not go to much details about terraform, but to have a basic idea

Provider.tf

provider "hcloud" {
    token = var.hcloud_token
}

Hetzner.tf

data "template_file" "userdata" {
  template = "${file("user-data.yml")}"
  vars = {
    hostname = var.domain
    sshdport = var.ssh_port
  }
}

resource "hcloud_server" "node" {
  count       = 5
  name        = "rke-${count.index}"
  image       = "ubuntu-18.04"
  server_type = "cx11"
  user_data   = data.template_file.userdata.rendered
}

Output.tf

output "IPv4" {
  value = hcloud_server.node.*.ipv4_address
}

In my user-data (cloud-init) template, the most important lines are these

  - usermod -a -G docker deploy
  - ufw allow 6443/tcp
  - ufw allow 2379/tcp
  - ufw allow 2380/tcp
  - ufw allow 80/tcp
  - ufw allow 443/tcp

build infra

$ terraform init
$ terraform plan
$ terraform apply

output

IPv4 = [
  "78.47.6x.yyy",
  "78.47.1x.yyy",
  "78.46.2x.yyy",
  "78.47.7x.yyy",
  "78.47.4x.yyy",
]

In the end we will see something like this on hetzner cloud

hetzner VMs

Rancher Kubernetes Engine

Take a look here for more details about what is required and important on using rke: Requirements.

We are going to use the rke aka the Rancher Kubernetes Engine, an extremely simple, lightning fast Kubernetes installer that works everywhere.

download

Download the latest binary from github:
Release Release v1.0.0

$ curl -sLO https://github.com/rancher/rke/releases/download/v1.0.0/rke_linux-amd64
$ chmod +x rke_linux-amd64
$ sudo mv rke_linux-amd64 /usr/local/bin/rke

version

$ rke --version

rke version v1.0.0

rke config

We are ready to configure our Kubernetes Infrastructure using the first 4 VMs.

$ rke config

master

[+] Cluster Level SSH Private Key Path [~/.ssh/id_rsa]:
[+] Number of Hosts [1]: 4
[+] SSH Address of host (1) [none]: 78.47.6x.yyy
[+] SSH Port of host (1) [22]:
[+] SSH Private Key Path of host (78.47.6x.yyy) [none]:
[-] You have entered empty SSH key path, trying fetch from SSH key parameter
[+] SSH Private Key of host (78.47.6x.yyy) [none]:
[-] You have entered empty SSH key, defaulting to cluster level SSH key: ~/.ssh/id_rsa
[+] SSH User of host (78.47.6x.yyy) [ubuntu]:
[+] Is host (78.47.6x.yyy) a Control Plane host (y/n)? [y]:
[+] Is host (78.47.6x.yyy) a Worker host (y/n)? [n]: n
[+] Is host (78.47.6x.yyy) an etcd host (y/n)? [n]: n
[+] Override Hostname of host (78.47.6x.yyy) [none]: rke-master
[+] Internal IP of host (78.47.6x.yyy) [none]:
[+] Docker socket path on host (78.47.6x.yyy) [/var/run/docker.sock]: 

etcd

[+] SSH Address of host (2) [none]: 78.47.1x.yyy
[+] SSH Port of host (2) [22]:
[+] SSH Private Key Path of host (78.47.1x.yyy) [none]:
[-] You have entered empty SSH key path, trying fetch from SSH key parameter
[+] SSH Private Key of host (78.47.1x.yyy) [none]:
[-] You have entered empty SSH key, defaulting to cluster level SSH key: ~/.ssh/id_rsa
[+] SSH User of host (78.47.1x.yyy) [ubuntu]:
[+] Is host (78.47.1x.yyy) a Control Plane host (y/n)? [y]: n
[+] Is host (78.47.1x.yyy) a Worker host (y/n)? [n]: n
[+] Is host (78.47.1x.yyy) an etcd host (y/n)? [n]: y
[+] Override Hostname of host (78.47.1x.yyy) [none]: rke-etcd
[+] Internal IP of host (78.47.1x.yyy) [none]:
[+] Docker socket path on host (78.47.1x.yyy) [/var/run/docker.sock]: 

workers

worker-01

[+] SSH Address of host (3) [none]: 78.46.2x.yyy
[+] SSH Port of host (3) [22]:
[+] SSH Private Key Path of host (78.46.2x.yyy) [none]:
[-] You have entered empty SSH key path, trying fetch from SSH key parameter
[+] SSH Private Key of host (78.46.2x.yyy) [none]:
[-] You have entered empty SSH key, defaulting to cluster level SSH key: ~/.ssh/id_rsa
[+] SSH User of host (78.46.2x.yyy) [ubuntu]:
[+] Is host (78.46.2x.yyy) a Control Plane host (y/n)? [y]: n
[+] Is host (78.46.2x.yyy) a Worker host (y/n)? [n]: y
[+] Is host (78.46.2x.yyy) an etcd host (y/n)? [n]: n
[+] Override Hostname of host (78.46.2x.yyy) [none]: rke-worker-01
[+] Internal IP of host (78.46.2x.yyy) [none]:
[+] Docker socket path on host (78.46.2x.yyy) [/var/run/docker.sock]: 

worker-02

[+] SSH Address of host (4) [none]: 78.47.4x.yyy
[+] SSH Port of host (4) [22]:
[+] SSH Private Key Path of host (78.47.4x.yyy) [none]:
[-] You have entered empty SSH key path, trying fetch from SSH key parameter
[+] SSH Private Key of host (78.47.4x.yyy) [none]:
[-] You have entered empty SSH key, defaulting to cluster level SSH key: ~/.ssh/id_rsa
[+] SSH User of host (78.47.4x.yyy) [ubuntu]:
[+] Is host (78.47.4x.yyy) a Control Plane host (y/n)? [y]: n
[+] Is host (78.47.4x.yyy) a Worker host (y/n)? [n]: y
[+] Is host (78.47.4x.yyy) an etcd host (y/n)? [n]: n
[+] Override Hostname of host (78.47.4x.yyy) [none]: rke-worker-02
[+] Internal IP of host (78.47.4x.yyy) [none]:
[+] Docker socket path on host (78.47.4x.yyy) [/var/run/docker.sock]: 

Network Plugin Type

[+] Network Plugin Type (flannel, calico, weave, canal) [canal]: 

rke_config

[+] Authentication Strategy [x509]:
[+] Authorization Mode (rbac, none) [rbac]: none
[+] Kubernetes Docker image [rancher/hyperkube:v1.16.3-rancher1]:
[+] Cluster domain [cluster.local]:
[+] Service Cluster IP Range [10.43.0.0/16]:
[+] Enable PodSecurityPolicy [n]:
[+] Cluster Network CIDR [10.42.0.0/16]:
[+] Cluster DNS Service IP [10.43.0.10]:
[+] Add addon manifest URLs or YAML files [no]: 

cluster.yml

the rke config will produce a cluster yaml file, for us to review or edit in case of misconfigure

$ ls -l cluster.yml
-rw-r----- 1 ebal ebal 4720 Dec  7 20:57 cluster.yml

rke up

We are ready to setup our KaaS by running:

$ rke up
INFO[0000] Running RKE version: v1.0.0
INFO[0000] Initiating Kubernetes cluster
INFO[0000] [dialer] Setup tunnel for host [78.47.6x.yyy]
INFO[0000] [dialer] Setup tunnel for host [78.47.1x.yyy]
INFO[0000] [dialer] Setup tunnel for host [78.46.2x.yyy]
INFO[0000] [dialer] Setup tunnel for host [78.47.7x.yyy]
...
INFO[0329] [dns] DNS provider coredns deployed successfully
INFO[0329] [addons] Setting up Metrics Server
INFO[0329] [addons] Saving ConfigMap for addon rke-metrics-addon to Kubernetes
INFO[0329] [addons] Successfully saved ConfigMap for addon rke-metrics-addon to Kubernetes
INFO[0329] [addons] Executing deploy job rke-metrics-addon
INFO[0335] [addons] Metrics Server deployed successfully
INFO[0335] [ingress] Setting up nginx ingress controller
INFO[0335] [addons] Saving ConfigMap for addon rke-ingress-controller to Kubernetes
INFO[0335] [addons] Successfully saved ConfigMap for addon rke-ingress-controller to Kubernetes
INFO[0335] [addons] Executing deploy job rke-ingress-controller
INFO[0341] [ingress] ingress controller nginx deployed successfully
INFO[0341] [addons] Setting up user addons
INFO[0341] [addons] no user addons defined
INFO[0341] Finished building Kubernetes cluster successfully 

Kubernetes

The output of rke will produce a local kube config cluster yaml file for us to connect to kubernetes cluster.

kube_config_cluster.yml

Let’s test our k8s !

$ kubectl --kubeconfig=kube_config_cluster.yml get nodes -A
NAME           STATUS   ROLES          AGE    VERSION
rke-etcd       Ready    etcd           2m5s   v1.16.3
rke-master     Ready    controlplane   2m6s   v1.16.3
rke-worker-1   Ready    worker         2m4s   v1.16.3
rke-worker-2   Ready    worker         2m2s   v1.16.3

$ kubectl --kubeconfig=kube_config_cluster.yml get pods -A
NAMESPACE       NAME                                      READY   STATUS      RESTARTS   AGE
ingress-nginx   default-http-backend-67cf578fc4-nlbb6     1/1     Running     0          96s
ingress-nginx   nginx-ingress-controller-7scft            1/1     Running     0          96s
ingress-nginx   nginx-ingress-controller-8bmmm            1/1     Running     0          96s
kube-system     canal-4x58t                               2/2     Running     0          114s
kube-system     canal-fbr2w                               2/2     Running     0          114s
kube-system     canal-lhz4x                               2/2     Running     1          114s
kube-system     canal-sffwm                               2/2     Running     0          114s
kube-system     coredns-57dc77df8f-9h648                  1/1     Running     0          24s
kube-system     coredns-57dc77df8f-pmtvk                  1/1     Running     0          107s
kube-system     coredns-autoscaler-7774bdbd85-qhs9g       1/1     Running     0          106s
kube-system     metrics-server-64f6dffb84-txglk           1/1     Running     0          101s
kube-system     rke-coredns-addon-deploy-job-9dhlx        0/1     Completed   0          110s
kube-system     rke-ingress-controller-deploy-job-jq679   0/1     Completed   0          98s
kube-system     rke-metrics-addon-deploy-job-nrpjm        0/1     Completed   0          104s
kube-system     rke-network-plugin-deploy-job-x7rt9       0/1     Completed   0          117s

$ kubectl --kubeconfig=kube_config_cluster.yml get componentstatus
NAME                 AGE
controller-manager   <unknown>
scheduler            <unknown>
etcd-0               <unknown>             <unknown>

$ kubectl --kubeconfig=kube_config_cluster.yml get deployments -A
NAMESPACE       NAME                   READY   UP-TO-DATE   AVAILABLE   AGE
ingress-nginx   default-http-backend   1/1     1            1           2m58s
kube-system     coredns                2/2     2            2           3m9s
kube-system     coredns-autoscaler     1/1     1            1           3m8s
kube-system     metrics-server         1/1     1            1           3m4s

$ kubectl --kubeconfig=kube_config_cluster.yml get ns
NAME              STATUS   AGE
default           Active   4m28s
ingress-nginx     Active   3m24s
kube-node-lease   Active   4m29s
kube-public       Active   4m29s
kube-system       Active   4m29s

Rancer2

Now login to the 5th VM we have in Hetzner:

ssh "78.47.4x.yyy" -l ubuntu -p zzzz

and install the stable version of Rancher2

$ docker run -d
    --restart=unless-stopped
    -p 80:80 -p 443:443
    --name rancher2
    -v /opt/rancher:/var/lib/rancher
    rancher/rancher:stable
    --acme-domain k8s.ipname.me

Caveat: I have create a domain and assigned to this hostname the IP of the latest VMs!
Now I can use letsencrypt with rancher via acme-domain.

verify

$ docker images -a
REPOSITORY          TAG                 IMAGE ID            CREATED             SIZE
rancher/rancher     stable              5ebba94410d8        10 days ago         654MB

$ docker ps -a -a
CONTAINER ID        IMAGE                    COMMAND                  CREATED             STATUS              PORTS                                      NAMES
8f798fb8184c        rancher/rancher:stable   "entrypoint.sh --acm…"   17 seconds ago      Up 15 seconds       0.0.0.0:80->80/tcp, 0.0.0.0:443->443/tcp   rancher2

Access

Before we continue, we need to give access to these VMs so they can communicate with each other. In cloud you can create a VPC with the correct security groups. But with VMs the easiest way is to do something like this:

sudo ufw allow from "78.47.6x.yyy",
sudo ufw allow from "78.47.1x.yyy",
sudo ufw allow from "78.46.2x.yyy",
sudo ufw allow from "78.47.7x.yyy",
sudo ufw allow from "78.47.4x.yyy",

Dashboard

Open your browser and type the IP of your rancher2 VM:

https://78.47.4x.yyy

or (in my case):

https://k8s.ipname.me

and follow the below instructions

rke_02.png

rke_03.png

rke_04.png

rke_05.png

rke_06.png

rke_07.png

Connect cluster with Rancher2

Download the racnher2 yaml file to your local directory:

$ curl -sLo rancher2.yaml https://k8s.ipname.me/v3/import/nk6p4mg9tzggqscrhh8bzbqdt4447fsffwfm8lms5ghr8r498lngtp.yaml

And apply this yaml file to your kubernetes cluster:

$ kubectl --kubeconfig=kube_config_cluster.yml apply -f rancher2.yaml

clusterrole.rbac.authorization.k8s.io/proxy-clusterrole-kubeapiserver unchanged
clusterrolebinding.rbac.authorization.k8s.io/proxy-role-binding-kubernetes-master unchanged
namespace/cattle-system unchanged
serviceaccount/cattle unchanged
clusterrolebinding.rbac.authorization.k8s.io/cattle-admin-binding unchanged
secret/cattle-credentials-2704c5f created
clusterrole.rbac.authorization.k8s.io/cattle-admin configured
deployment.apps/cattle-cluster-agent configured
daemonset.apps/cattle-node-agent configured

Web Gui

rke_08.png

rke_09.png

kubectl config

We can now use the Rancher kubectl config by downloading from here:

rke_09b.png

In this post, it is rancher2.config.yml

helm

Final step is to use helm to install an application to our kubernetes cluster

download and install

$ curl -sfL https://get.helm.sh/helm-v3.0.1-linux-amd64.tar.gz | tar -zxf -

$ chmod +x linux-amd64/helm
$ sudo mv linux-amd64/helm /usr/local/bin/

Add Repo

$ helm repo add stable https://kubernetes-charts.storage.googleapis.com/
"stable" has been added to your repositories

$ helm repo update
Hang tight while we grab the latest from your chart repositories...
...
Successfully got an update from the "stable" chart repository
Update Complete. ⎈ Happy Helming!⎈ 

weave-scope

Install weave scope to rancher:

$ helm --kubeconfig rancher2.config.yml install stable/weave-scope --generate-name
NAME: weave-scope-1575800948
LAST DEPLOYED: Sun Dec  8 12:29:12 2019
NAMESPACE: default
STATUS: deployed
REVISION: 1
NOTES:
You should now be able to access the Scope frontend in your web browser, by
using kubectl port-forward:

kubectl -n default port-forward $(kubectl -n default get endpoints
weave-scope-1575800948-weave-scope -o jsonpath='{.subsets[0].addresses[0].targetRef.name}') 8080:4040

then browsing to http://localhost:8080/.
For more details on using Weave Scope, see the Weave Scope documentation:

https://www.weave.works/docs/scope/latest/introducing/

Proxy

Last, we are going to use kubectl to create a forwarder

$ kubectl --kubeconfig=rancher2.config.yml -n default port-forward $(kubectl --kubeconfig=rancher2.config.yml -n default get endpoints weave-scope-1575800948-weave-scope -o jsonpath='{.subsets[0].addresses[0].targetRef.name}') 8080:4040
Forwarding from 127.0.0.1:8080 -> 4040
Forwarding from [::1]:8080 -> 4040

Open your browser in this url:

  http://localhost:8080

rke_10.png

That’s it !