Kuberenetes Upgrade Process

 

[Kubernetes Release]

 

https://kubernetes.io/docs/concepts/overview/kubernetes-api/

The Kubernetes API

Here is a link to kubernetes documentation if you want to learn more about this topic (You don't need it for the exam though):

https://github.com/kubernetes/community/blob/master/contributors/devel/sig-architecture/api-conventions.md

kubernetes/community

https://github.com/kubernetes/community/blob/master/contributors/devel/sig-architecture/api_changes.md

kubernetes/community

[Cluster Upgrade Process]

 

Kubernets에서는 kube-apiserver보다 동일 또는 하위 버전이 되어야 한다.

업그레이드는 마이너버전 한 단계씩..

 

 

kubectl get pods -A

node 버전 확인

kubectl get nodes

 

Upgrade를 위한 버전 결정

 

(Ubuntu, Debian or HypriotOS)

apt update
apt-cache madison kubeadm

(CentOS, RHEL or Fedora)

yum list --showduplicates kubeadm --disableexcludes=kubernetes

 

Upgrade Control plane nodes

 

 

[Practice Test]

ㅁ What is the current version of the cluster?

kubectl get nodes

NAME           STATUS   ROLES    AGE     VERSION
controlplane   Ready    master   5m25s   v1.18.0
node01         Ready    <none>   4m57s   v1.18.0

kubectl version --short

 

ㅁ How many nodes are part of this cluster?

[0], [2], [3], [1], [4]

 

ㅁ How many nodes can host workload in this cluster (Inspect the applications and taints set on the nodes)

- Check the taints on both master and node01. If none exists, then both nodes can host workloads.

[4], [2], [3], [1]

kubectl describe node controlplane | grep -i Taint
Taints:             <none>

kubectl describe node node01 | grep -i Taint
Taints:             <none>

ㅁ How many applications are hosted on the cluster?

kubectl get deployment

ㅁ What nodes are the pods hosted on?

kubectl get pods -o wide

ㅁ You are tasked to upgrade the cluster. User's accessing the applications must not be impacted. And you cannot provision new VMs. What strategy would you use to upgrade the cluster?

[Upgrade one node at a time while moving the workloads to the other]

[Users will be impacted since there is only one worker node]

[Add new nodes with newer versions while taking down existing nodes]

[Upgrade all nodes at once] 

 

ㅁ What is the latest stable version available for upgrade?

 

[v1.12.1]

[v1.13.0]

[v1.18.17]

[v1.10.0]

kubeadm upgrade plan

[upgrade/config] Making sure the configuration is correct:
[upgrade/config] Reading configuration from the cluster...
[upgrade/config] FYI: You can look at this config file with 'kubectl -n kube-system get cm kubeadm-config -oyaml'
[preflight] Running pre-flight checks.
[upgrade] Running cluster health checks
[upgrade] Fetching available versions to upgrade to
[upgrade/versions] Cluster version: v1.18.0
[upgrade/versions] kubeadm version: v1.18.0
I0328 04:08:40.186928   32456 version.go:252] remote version is much newer: v1.20.5; falling back to: stable-1.18
[upgrade/versions] Latest stable version: v1.18.17
[upgrade/versions] Latest stable version: v1.18.17
[upgrade/versions] Latest version in the v1.18 series: v1.18.17
[upgrade/versions] Latest version in the v1.18 series: v1.18.17

Components that must be upgraded manually after you have upgraded the control plane with 'kubeadm upgrade apply':
COMPONENT   CURRENT       AVAILABLE
Kubelet     2 x v1.18.0   v1.18.17

Upgrade to the latest version in the v1.18 series:

COMPONENT            CURRENT   AVAILABLE
API Server           v1.18.0   v1.18.17
Controller Manager   v1.18.0   v1.18.17
Scheduler            v1.18.0   v1.18.17
Kube Proxy           v1.18.0   v1.18.17
CoreDNS              1.6.7     1.6.7
Etcd                 3.4.3     3.4.3-0

You can now apply the upgrade by executing the following command:

        kubeadm upgrade apply v1.18.17

Note: Before you can perform this upgrade, you have to update kubeadm to v1.18.17.

_____________________________________________________________________

 

ㅁ We will be upgrading the master node first. Drain the master node of w orkloads and mark it

kubectl drain controlplane --ignore-deamonsets

node/controlplane cordoned
WARNING: ignoring DaemonSet-managed Pods: kube-system/kube-flannel-ds-amd64-6bh67, kube-system/kube-keepalived-vip-d55fx, kube-system/kube-proxy-b9vm4
evicting pod kube-system/coredns-66bff467f8-2qzhx
evicting pod kube-system/coredns-66bff467f8-gczlb
pod/coredns-66bff467f8-gczlb evicted
pod/coredns-66bff467f8-2qzhx evicted
node/controlplane evicted

kubectl get nodes
NAME           STATUS                     ROLES    AGE   VERSION
controlplane   Ready,SchedulingDisabled   master   21m   v1.18.0
node01         Ready                      <none>   21m   v1.18.0

 

ㅁ Upgrade the master/controlplane compnents to exact version v1.19.0

Upgrade kubeadm tool (if not already), then the master components, and fnally the kubelet. Practice referring to the kubernetes documentation page. Note: while upgrading kubelet, if you hit dependency issue while running the apt-get upgrade kubelet command, use the apt install kubelet=1.19.0-00 command instead

 

ㅇ Master Node

 

1) 먼저 kubeadm upgrade plan을 수행하여 업그레이드 가능 버전 확인

kubeadm upgrade plan

1.18.17-00 으로 확인되어 kubeadm 업데이트

 

2) apt install kubeadm=1.18.17-00 업데이트

apt install kubeadm=1.18.17-00

업데이트 후 kubeadm 버전 확인

kubeadm version

3) master node의 업데이트

 

먼저 master node 상태 확인

kubectl get nodes

 

master node upgrade

kubeadm upgrade apply v1.18.17

 

kubelet upgrade

apt install kubelet=1.18.17-00

 

업데이트 여부 확인

kubectl get nodes

 

4) 정상 상태(Schedulable)로 Controlplane 변경

 

ㅇ Worker Node Udpate

1) Worker Node 상태 drain 으로 변경

kubectl drain node01 --ignore-daemonsets

 

2) Worker Node로 접속

ssh node01

 

 

3) kubeadm update 

 

kubeadm 버전 확인

kubeadm version

kubeadm upgrade

apt install kubeadm=1.18.17-00

 

4) worker node kubelet update

apt install kubelet=1.18.17-00

 

5) Control Plane으로 접속하여 Uncordon 함

kubectl uncordon node01

 

ㅇ 최종적으로 모두 업데이트 되었는지 확인

kubectl get nodes

kubectl version --short