3.1 Installation of kubernetes on debian11 and Ubuntu 22.0.4 in Vmware

rafal.radziejewski

Some debug info attached in the screen. Can anybody help me to debug what's wrong ?

The connection to the server k8scp:6443 was refused - did you specify the right host or port?

rafal.radziejewski

rafal.radziejewski

rafal.radziejewski

From time to time api-server is available on port 6443 but mostly not

Seems like api-server was looking for etcd which is not available.

Logs from etcd:

rafal.radziejewski

Worth to mention it is vmware workstation 16.1

Here are logs on api-server somehow it is shutting down for no reason.

Same for etcd

Same for scheduler

Seems like kubelet is shutting them down for some reason ?

rafal.radziejewski

Some logs from kubelet

Can somebody please help ?

chrispokorni

Hi @rafal.radziejewski,

Thanks for all the detailed screenshots.

Please revisit the lab guide for the recommended Ubuntu version, that is 20.04 LTS. Version 22.04 LTS has not yet been tested with the lab exercises, and there may be dependencies that still need to be worked out. Other learners had similar experiences on 22, and eventually rebuilt their environments successfully on Ubuntu 20.

I would recommend watching the two demo videos from the introductory chapter on setting up the infrastructure for labs. They are referencing GCP and AWS cloud infrastructure, however, when setting up the VM to VM networking, they both provide key networking options that need to be properly set in order for the cluster nodes to communicate with each other. In this particular scenario, those networking options will need to be set from the VMware hypervisor to resemble the ones presented for the cloud.

Also, please ensure the recommended amount of CPU, memory and disk space are assigned to each VM instance. Insufficient compute resources on a VM/node may cause kubelet and/or other control plane components to panic and unexpectedly terminate. This also means that your host machine should be powerful enough to support its own OS, host applications, the hypervisor, and the minimum of two VMs for the lab environment.

Regards,
-Chris

rafal.radziejewski

Thank you @chrispokorni i will try 20.04 and come back there

headkaze

I had no problem with 22.04 LTS but I did have to increase the CPU/memory/storage before it worked.

So for control-plane's I used 2 VCPU, 2GB RAM, 30GB Storage and for my workers I used 1 VCPU, 2GB RAM, 30GB Storage.

It's probably a bit of an overkill though!

rafal.radziejewski

I was using 4VCPUS, 8GB RAM, 50GB Storage for control plane with 22.04 and i did reinstall 5 times, no success.
hri
With 20.04 everything works like a charm thanks once again . @chrispokorni