Description

This quick setup guide helps in setting up a kubernetes cluster on Ubuntu system which can later be used to deploy NetIQ Access Manager Docker images (Beta).

Detailed Description

In this example we will create 2 node Kubernetes cluster.

These will be self managed nodes and any updates to Virtual Machine or Kubernetes will need to be managed by the administrator creating the nodes.

Lets start by creating 2 Ubuntu Virtual Machines, lets name them as kube-master and other as kubenode-1.

We can have more than 2 virtual machine too and the consecutive Virtual Machines will added as nodes/worker nodes.

In the below approach, we will adopt the kubeadm feature of kubernetes to setup the cluster and use Flannel Networking feature.

Prerequisites

Create 2 Ubuntu VMs and setup necessary networking/IP etc and Set Unique hostname (/etc/hostname) and then Update /etc/hosts on both the boxes.

Disable swap  by executing "swapoff -a" on both the boxes.

Reboot the boxes.

Example uses "Ubuntu 18.04.1 LTS".

High-level Tasks:

1) Installing Binaries on all the nodes

2) Configuring Kubernetes on master node

3) Join Nodes to Kubernetes master

4) Deploy NetIQ Access Manager Docker images using helm

Part 1 - Installing the binaries:

Note: The below needs to be executed on all the nodes.

Step 1:

Install Docker Engine:

apt-get update

apt-get install -y docker.io

Step 2:

Install Kubernetes binaries:

sudo apt-get update && sudo apt-get install -y apt-transport-https curl

curl -s https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -

cat <<EOF | sudo tee /etc/apt/sources.list.d/kubernetes.list
deb https://apt.kubernetes.io/ kubernetes-xenial main
EOF

sudo apt-get update
sudo apt-get install -y kubelet kubeadm kubectl

(Optional)sudo apt-mark hold kubelet kubeadm kubectl

Enable docker service:

systemctl enable docker.service

Now, the required binaries are installed.

Part 2- Configuring Kubernetes cluster - Master Node:

Below section applies to Master Node:

NetIQ Access Manager deploys the image using Helm tool.

It's an easy to use and manager kubernetes yaml files and acts as a wrapper.

Here we are installing Helm version 3.0 on the Master node by executing the below command.

Install Helm v 3.0:

curl https://raw.githubusercontent.com/helm/helm/master/scripts/get-helm-3 | bash

Execute the below commands on Node which needs to be designated as master.

***************************************

swapoff -a

(kubelet, a kubernetes agent software doesn't work when swap is enabled

for more information refer:
https://github.com/kubernetes/kubernetes/issues/53533)

kubeadm init --pod-network-cidr=10.244.0.0/16

*** (record the output of the above command

Output displays a join command needs to be executed later to add nodes to kubernetes cluster)

sysctl net.bridge.bridge-nf-call-iptables=1

export KUBECONFIG=/etc/kubernetes/admin.conf

I chose Flannel network:

kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml

***************************************

Now that the Master is configured, let's verify if things are up on the Master Node.

Verification:

kubectl get pods --all-namespaces

Terminal Output would appear as below

NAMESPACE     NAME                                  READY   STATUS    RESTARTS   AGE

kube-system   coredns-5644d7b6d9-9kcdp              1/1     Running   0          4m26s

kube-system   coredns-5644d7b6d9-pt2g9              1/1     Running   0          4m26s

kube-system   etcd-kube-master                      1/1     Running   0          3m46s

kube-system   kube-apiserver-kube-master            1/1     Running   0          3m26s

kube-system   kube-controller-manager-kube-master   1/1     Running   0          3m46s

kube-system   kube-flannel-ds-amd64-h9dgq           1/1     Running   0          33s

kube-system   kube-proxy-2vjdt                      1/1     Running   0          4m26s

kube-system   kube-scheduler-kube-master            1/1     Running   0          3m45s

***************************************

Part 3- Adding Nodes to Kubernetes Cluster

Below section applies to Worker Node:

Execute the below command on all Nodes which needs to be designated as worker node.

(the token will differ and the below is just an example)

***************************************

kubeadm join 10.71.131.170:6443 --token 0y9jwq.zkixei59b6r7rbh8 --discovery-token-ca-cert-hash sha256:7b93e15d454089885fc3ec12a832579dda9a84171e944ff2b745075da38ebc71

***************************************

The cluster is configured and it's up and running.

Let's verify if things are as expected.

Execute the below on Master Node to verify the cluster status.

kubectl get nodes

Output should be similar to below:

NAME          STATUS   ROLES    AGE     VERSION

kube-master   Ready    master   7m58s   v1.16.3

kubenode1     Ready    <none>   2m27s   v1.16.3

Part 4- Deploying NetIQ Access Manager Docker images (beta) to Kubernetes

Now, follow along the NetIQ Access Manager docker deployment documentation to deploy NAM docker images to Kubernetes Cluster.

Refer to: 

beta release https://community.microfocus.com/t5/Beta-Release-of-NetIQ-Access/Announcement-Beta-Release-of-deploying-Access-Manager-in-Docker/m-p/2832015#M1