Setting up Kubernetes cluster on FABRIC

[Acheme Acheme]

Hello,

I will be very grateful if someone can share a working notebook or steps to successfully setup a kubernetes cluster (a master node and at least one worker node) on FABRIC.

The example notebook configuration seems to be out of date and I tried to setup using guides online with issues:

After installing the application on the Master node and two worker nodes, I install a Container Network Interface (CNI) application and initialize the on the master.
I have able to “kubectl get nodes” and see the master node is “ready”. but using the join message on the worker nodes gives an error:ubuntu@Node3:~$ sudo kubeadm join 10.20.4.87:6443 –token 328ikr.ix01rv2tss194wpm –discovery-token-ca-cert-hash sha256:f4f2be6d0e13d8a30f660fd22836f85b7a6c6068115910dc8a73634eef197540
[preflight] Running pre-flight checks
error execution phase preflight: couldn’t validate the identity of the API Server: failed to request the cluster-info ConfigMap: client rate limiter Wait returned an error: rate: Wait(n=1) would exceed context deadline
To see the stack trace of this error execute with –v=5 or higher

Thank you very much.

[Fraida Fund]

Hi, you can use this example: https://github.com/teaching-on-testbeds/k8s

I just tested it and the playbook failed on the first run, but was successful on a second attempt. If it is successful, you should see zero “failed”, like this –

PLAY RECAP *********************************************************************
localhost                  : ok=3    changed=0    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0   
node-0                     : ok=715  changed=35   unreachable=0    failed=0    skipped=1252 rescued=0    ignored=1   
node-1                     : ok=610  changed=26   unreachable=0    failed=0    skipped=1109 rescued=0    ignored=1   
node-2                     : ok=502  changed=20   unreachable=0    failed=0    skipped=779  rescued=0    ignored=1   

(I developed that example for teaching this material: https://github.com/teaching-on-testbeds/k8s-ml, if you want to see an example of how it is used. )

[Acheme Acheme]

Thank you very much, this works!

[Komal Thareja]

Thank you so much @Fraida! Could we please request you to consider uploading this to Fabric Artifacts to enable other Fabric users to leverage this?

Appreciate your help with this!

Artifact Manager: https://artifacts.fabric-testbed.net/artifacts/

[Prabhudatta Mishra]

_*]:min-w-0 !gap-3.5″>

I’m struggling to get any CNI (Flannel, Weave, Calico) working on my 3-node Kubernetes cluster on FABRIC testbed. The. master node is Ready, core components are running, but all CNI solutions consistently fail with CrashLoopBackOff. CoreDNS pods remain stuck in ContainerCreating state.

The nodes are connected via L2 network (interface enp7s0, IPs 192.168.100.10-12) using containerd 1.7.27 with K8s v1.32.5.

Has anyone successfully deployed a working CNI on FABRIC? Any guidance on resolving Docker-Kubernetes iptables conflicts, interface detection issues, or CNI-specific configurations for FABRIC would be greatly appreciated.

[Acheme Acheme]

Hello Prabhudatta,

I used the resource in the github (https://github.com/teaching-on-testbeds/k8s/blob/main/reserve_fabric/index.md) provided by Fraida Fund which installs Calico CNI.

You will need to clone repo: https://github.com/teaching-on-testbeds/k8s, and follow the instructions in index.md.

I have shared the notebook from same repo which I used for output reference.

1 user thanked author for this post.

Prabhudatta Mishra

[Author]

Posts