LPS:Kubernetes/Upgrade

tato strankapopisuje upgrade kubernetes nodu

system

master

vypsat nody se systemem a verzi z master nodu

kubectl get no -o wide

vysusit node

kubectl drain <node> --ignore-daemonsets

zkontrolovat, ze uz je SchedulingDisabled

kubectl get no

node

• na upgradovanem nodu se udela simpleupgrade https://ipmil.civ.zcu.cz/linux/upgraduj#simpleupgrade
• aplikovat salt predpisy pred rebootem

salt-call state.apply
apt upgrade -y

kdyz ti zlobi file.managed z url, muze to byt problem salt cache. soubor se pak z url nerefreshuje, ale pouzije se nacachovana kopie

salt-call saltutil.clear_cache

• po rebootu

cf-agent -K; apt update -qq; apt upgrade -y

master

kdy je vsechno ok, vrat node do provozu koukej, jeste jede ok (Ready)

kubectl uncordon <node>
kubectl get no

kubernetes

projed si change log

https://github.com/kubernetes/kubernetes/blob/master/CHANGELOG/CHANGELOG-1.23.md

a pak jedeme podle navodu

https://kubernetes.io/docs/tasks/administer-cluster/kubeadm/kubeadm-upgrade/

# master node
apt-mark unhold kubeadm cri-tools
apt-get update -qq && apt-get install -y kubeadm=1.22.11-00
apt-mark hold kubeadm cri-tools

# only on master node
# kubeadm upgrade plan
# kubeadm upgrade apply v1.22.11

# For the other control plane nodes
kubeadm upgrade node

# Upgrade kubelet and kubectl

kubectl drain --ignore-daemonsets synergia

apt-mark unhold kubelet kubectl
apt-get install -y kubelet=1.22.11-00 kubectl=1.22.11-00
apt-mark hold kubelet kubectl
systemctl daemon-reload
systemctl restart kubelet

kubectl uncordon synergia
kubectl get nodes