Welcome to the Linux Foundation Forum!

Issue in setting up local lab env (Lab 2.2)

andreacervo
andreacervo Posts: 4
in LFD259 Class Forum

Hello,
I'm writing to you because I'm having a problem following Lab 2.2 of the LFD259 course.
I'm setting up a local lab with a Windows 11 host and a VirtualBox hypervisor.
The control plane and worker VMs are based on the ISO ubuntu-24.04.3-live-server-amd64.iso.

I'm using this tarball version: LFD259_V2025-05-28_SOLUTIONS.tar.xz

I'm stuck at step 6 on page 2, when I have to execute the command "bash k8scp.sh | tee $HOME/cp.out"

I get this error in output: "Detected that the sandbox image "registry.k8s.io/pause:3.8" of the container Runtime is inconsistent with that used by kubeadm. It is recommended to use "registry.k8s.io/pause:3.10" as the CRI sandbox image."

How should I proceed?

Thank you in advance for your support, best regards
Andrea

Answers

  • chrispokorni
    chrispokorni Posts: 2,501

    Hi @andreacervo,

    For a while now that message has been consistently showing during the control plane init phase. Feel free to disregard it and continue with the exercise.

    Regards,
    -Chris

  • andreacervo
    andreacervo Posts: 4

    Thanks @chrispokorni ,
    what should I do in practice to disregard it?
    Does this mean the installation was successful anyway? What should I check to ensure the next steps meet the necessary prerequisites?

    Thanks, kind regards
    Andrea

  • chrispokorni
    chrispokorni Posts: 2,501

    Hi @andreacervo,

    You can simply compare your output with the sample output from the lab guide. Scan through your output, or the cp.out file and try to locate towards the end "Your Kubernetes control-plane has initialized successfully!" and a long multi-line kubeadm join ... command (followed by your control plane VM private IP, a unique token, a unique sha256 hash), and "Continue to the next step". If you find them it means your control plane is configured and operational, so you can proceed to the next step in the exercise.

    Regards,
    -Chris

  • andreacervo
    andreacervo Posts: 4

    Ciao @chrispokorni,
    Unfortunately, running a command like grep "Kubernetes control-plane" cp.out I can't find anything in the cp.out document.

    The last lines of the document are these strings:

    crictl
    [init] Using Kubernetes version: v1.33.1
    [preflight] Running pre-flight checks
    [preflight] Pulling images required for setting up a Kubernetes cluster
    [preflight] This might take a minute or two, depending on the speed of your internet connection
    [preflight] You can also perform this action beforehand using 'kubeadm config images pull'

    How should I proceed? Should I try with the previous tarball version?

    Thanks, BR
    Andrea

  • chrispokorni
    chrispokorni Posts: 2,501

    Hi @andreacervo,

    Did you happen to execute the k8scp.sh script several times on the control plane VM?

    Also, please provide details about your cluster VMs to make sure they meet the requirements:

    • cloud (which provide) or local (which hypervisor, which arch: intel/amd or arm),
    • size (CPU, RAM, vDisk (dynamic allocation or full pre-allocation) ),
    • count of network interfaces, type of private network (host, nat, bridged),
    • guest OS,
    • private IP address,
    • firewalls.

    Regards,
    -Chris

  • andreacervo
    andreacervo Posts: 4

    Hi @chrispokorni
    I executed the script only once on this specific VM , but previously trying on a different VM with a different Ubuntu distro I obtained the same result.

    Here some details about the current setup:

    • local environment (no cloud), based on a host with intel core i5-1235U CPU and Windows 11 OS
    • VirtualBox 7.2
    • VM OS: ubuntu-24.04.3-live-server-amd64.iso
    • ControlPlane VM specifications: 3 CPUs, 6132 MB of RAM, 25 GB disk (full pre-allocation)
    • Network interface via bridged adapter (unique interface)
    • controlplane IP address (with bridged adapter): 192.168.1.160
    • firewall and swap deactivated, apparmor as well

    Thanks for the support, BR
    Andrea

  • chrispokorni
    chrispokorni Posts: 2,501

    Hi @andreacervo,

    From what I can tell the VM appears to satisfy the Kubernetes cluster requirements. I would suggest to ensure that for each VirtualBox VM the bridged adapter has the promiscuous mode enabled and "allow all" is selected for inbound traffic (from the VirtualBox console, stop the VM to make this change, then start it again). Make sure the hardware virtualization option and nested virtualization option are selected in VirtualBox for each VM. And, perhaps, increase the control plane VM's RAM to about 8 GB (from the current 6 GB) - I have seen Kubernetes perform better with more RAM than less :wink:, particularly with the Cilium plugin.

    On your Windows 11 host, disable Hyper-V if it was previously enabled - as it may conflict with VBox, and ensure the Intel virtualization feature is enabled in BIOS.

    Otherwise you should be good to give it another try. Disregard the sandbox image version mismatch, and be patient with the preflight checks - they could take about 4-5 minutes. Allow the k8scp.sh script to complete before continuing with the next step.

    Regards,
    -Chris

Categories

Upcoming Training