Can not set hostNetwork for static Pod to false. This results all static pods get same IP Address, which is misleading

[rakarakhi]

What keywords did you search in kubeadm issues before filing this one?

If you have found any duplicates, you should instead reply there and close this page.

If you have not found any duplicates, delete this section and continue on.

Is this a BUG REPORT or FEATURE REQUEST? BUG REPORT

Choose one: BUG REPORT or FEATURE REQUEST

Versions

kubeadm version (use kubeadm version):

kubeadm version: &version.Info{Major:"1", Minor:"22", GitVersion:"v1.22.6", GitCommit:"f59f5c2fda36e4036b49ec027e556a15456108f0", GitTreeState:"clean", BuildDate:"2022-01-19T17:31:49Z", GoVersion:"go1.16.12", Compiler:"gc", Platform:"linux/amd64"}

Environment:

• Kubernetes version (use kubectl version):

Client Version: version.Info{Major:"1", Minor:"22", GitVersion:"v1.22.6", GitCommit:"f59f5c2fda36e4036b49ec027e556a15456108f0", GitTreeState:"clean", BuildDate:"2022-01-19T17:33:06Z", GoVersion:"go1.16.12", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"22", GitVersion:"v1.22.6", GitCommit:"f59f5c2fda36e4036b49ec027e556a15456108f0", GitTreeState:"clean", BuildDate:"2022-01-19T17:26:47Z", GoVersion:"go1.16.12", Compiler:"gc", Platform:"linux/amd64"}

• Cloud provider or hardware configuration: AWS EC2 Instance: t3a-medium (2 vCPU and 4 GB RAM)
• OS (e.g. from /etc/os-release):

NAME="Ubuntu"
VERSION="18.04.6 LTS (Bionic Beaver)"
ID=ubuntu
ID_LIKE=debian
PRETTY_NAME="Ubuntu 18.04.6 LTS"
VERSION_ID="18.04"
HOME_URL="https://www.ubuntu.com/"
SUPPORT_URL="https://help.ubuntu.com/"
BUG_REPORT_URL="https://bugs.launchpad.net/ubuntu/"
PRIVACY_POLICY_URL="https://www.ubuntu.com/legal/terms-and-policies/privacy-policy"
VERSION_CODENAME=bionic
UBUNTU_CODENAME=bionic

• Kernel (e.g. uname -a): 5.4.0-1065-aws #68~18.04.1-Ubuntu
• Container runtime (CRI) (e.g. containerd, cri-o): containerd
• Container networking plugin (CNI) (e.g. Calico, Cilium): Calico
• Others: Ubuntu 18.04.6 LTS

What happened?

• In kubeadm init I am setting the pod CIDR. But all static pods have the same IP Address as hostNetwork=true.
• Changing that to false does not restart the pods with unique IP Addresses
• The entire manifest folder was missing
• Restoring the manifest folder from backup copy does not help

What do you expect to happen?

• Since I am setting the pod CIDR, all static pods should take unique IP Addresses.

How to reproduce it (as minimally and precisely as possible)?

• In kubeadm init pass the pod CIDR
• Install a pod CNI - Calico/Flannel
• kubectl get pods -n kube-system -o wide

Anything else we need to know?

[neolit123]

kubeadm deploys the control plane components with hostNetwork: true because they may have to interact the host network interfaces. you can use kubeadm patches to disable the hostNetwork option in the pod spec of the static pods during kubeadm init and join:
https://kubernetes.io/docs/setup/production-environment/tools/kubeadm/control-plane-flags/#patches

but note that having hostNetwork: false is not a supported use case by kubeadm...

Changing that to false does not restart the pods with unique IP Addresses

i don't think changing the podspec...hostNetwork of a static pod will result in good behavior after the components are already running... if you are seeing bugs around that you can try logging an issue in kubernetes/kubernetes because the kubelet manages the static pod restarts, not kubeadm.

The entire manifest folder was missing.

this should not happen. the kubelet does not delete files there...

Restoring the manifest folder from backup copy does not help.

just a speculation on my side, but it could be that after hostNetwork is changed to false, the kube-apiserver fails to start and the kubelet starts complaining about not being able to register the mirror pods.

either way, i don't see a kubeadm bug here, since this is not a supported scenario.
you might have to start the node from scratch.

/support

[github-actions]

Hello, @rakarakhi 🤖 👋

You seem to have troubles using Kubernetes and kubeadm.
Note that our issue trackers should not be used for providing support to users.
There are special channels for that purpose.

Please see:

• https://github.com/kubernetes/kubeadm#support

[rakarakhi]

Hello Ivanov, Hope I have addressed you by your first name. Many thanks for reaching out. I agree the pods need to network. And when I am using a set of VMs - either by Virtual Box on my local machine or in Cloud - it is already placed in the subnet. Yes, I need to deploy a CNI like calico or flannel to enable pod networking. The issue is as per theory pods too do get specific IP, which in this case is a direct contrast. I will check that link to see the option. And implement to see how it works. but for UX the present representation is misleading. Thanks a lot for your time and support. Regards, *Rahuldeb Chakrabarty* Email: ***@***.*** https://www.linkedin.com/in/RahuldebChakrabartyIITB

…

On Mon, Feb 14, 2022 at 9:02 PM Lubomir I. Ivanov ***@***.***> wrote: kubeadm deploys the control plane components with hostNetwork: true because they may have to interact the host network interfaces. you can use kubeadm patches to disable the hostNetwork option in the pod spec of the static pods during kubeadm init and join: https://kubernetes.io/docs/setup/production-environment/tools/kubeadm/control-plane-flags/#patches but note that having hostNetwork: false is not a supported use case by kubeadm... Changing that to false does not restart the pods with unique IP Addresses i don't think changing the podspec...hostNetwork of a static pod will result in good behavior after the components are already running... if you are seeing bugs around that you can try logging an issue in kubernetes/kubernetes because the kubelet manages the static pod restarts, not kubeadm. The entire manifest folder was missing. this should not happen. the kubelet does not delete files there... Restoring the manifest folder from backup copy does not help. just a speculation on my side, but it could be that after hostNetwork is changed to false, the kube-apiserver fails to start and the kubelet starts complaining about not being able to register the mirror pods. either way, i don't see a kubeadm bug here, since this is not a supported scenario. you might have to start the node from scratch. /support — Reply to this email directly, view it on GitHub <#2655 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGZLL42LHWCYIPLBTHGPOL3U3EM2JANCNFSM5OJBIT4Q> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>