Dear Kubernetes team and users,
I have recently got 16 nodes of a cluster that was going to be recycled. I am willing to use it to perform high performance calculations and for this I want to distribute jobs on different nodes.
I have understood that kubernetes can perform this but I have some questions:
- Do you have an easy installation tutorial which explains everything from scratch? For instance, I do not know how my network needs to be setup, how to install and configure kubernetes, what should I do to restart kubernetes in the case the cluster needs to be restarted.
- With Kubernetes, is it possible to start a job on multiple nodes? My understanding is that we can have multiple pods by nodes but not one pod on multiple nodes. Hence I am not sure I will be able to perform calculations on 128 cpu for instance. I have to say I have no knowledge of MPI and I do not even know if my calculation programs are compatible with.
Thank you very much for your help, it will be my first setup of a cluster so, please explain me as a 12 years old kid
Best,
Froz