kubernetes
/
website
mirror of https://github.com/kubernetes/website.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

docs(kubeadm-upgrade): add consideration on etcd upgrade impact 

relates to https://github.com/kubernetes/kubeadm/issues/2991#issuecomment-1932337556

Co-authored-by: Lubomir I. Ivanov <neolit123@gmail.com>

Signed-off-by: Clément Nussbaumer <clement.nussbaumer@postfinance.ch>
pull/45095/head
Clément Nussbaumer 2024-02-12 08:18:31 +01:00
parent fdc49a6ba8
commit 2dc571df77
No known key found for this signature in database
GPG Key ID: 5242D041B556BA42
1 changed files with 19 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
content/en/docs/tasks/administer-cluster/kubeadm/kubeadm-upgrade.md
View File

@ -54,11 +54,28 @@ The upgrade workflow at high level is the following:
with the purpose of reconfiguring the cluster is not recommended and can have unexpected results. Follow the steps in
[Reconfiguring a kubeadm cluster](/docs/tasks/administer-cluster/kubeadm/kubeadm-reconfigure) instead.
### Considerations when upgrading etcd
Because the `kube-apiserver` static pod is running at all times (even if you
have drained the node), when you perform a kubeadm upgrade which includes an
etcd upgrade, in-flight requests to the server will stall while the new etcd
static pod is restarting. As a workaround, it is possible to actively stop the
`kube-apiserver` process a few seconds before starting the `kubeadm upgrade
apply` command. This permits to complete in-flight requests and close existing
connections, and minimizes the consequence of the etcd downtime. This can be
done as follows on control plane nodes:
```shell
killall -s SIGTERM kube-apiserver # trigger a graceful kube-apiserver shutdown
sleep 20 # wait a little bit to permit completing in-flight requests
kubeadm upgrade ... # execute a kubeadm upgrade command
```
<!-- steps -->
## Changing the package repository
If you're using the community-owned package repositories (`pkgs.k8s.io`), you need to
If you're using the community-owned package repositories (`pkgs.k8s.io`), you need to
enable the package repository for the desired Kubernetes minor release. This is explained in
[Changing the Kubernetes package repository](/docs/tasks/administer-cluster/kubeadm/change-package-repository/)
document.
@ -320,4 +337,4 @@ and post-upgrade manifest file for a certain component, a backup file for it wil
`kubeadm upgrade node` does the following on worker nodes:
- Fetches the kubeadm `ClusterConfiguration` from the cluster.
- Upgrades the kubelet configuration for this node.
- Upgrades the kubelet configuration for this node.