Re: [kubernetes-users] Shared Kubernetes Control Plane

On Sat, Oct 1, 2016 at 6:05 PM, Mike wrote: > Thank you, Tim, for the reply. I sure better understand your logic. Given I > am new to Kubernetes, would you please shed some lights on where the > technical challenges are for such architecture (shared control plane)? Off the top of my head: * Maki

Re: [kubernetes-users] Shared Kubernetes Control Plane

There's a discussion here that is related; though it assumes only one worker group (using your terminology) per node. I tend to agree with what Daniel and Tim are saying. If you were willing to drop the requirement "groups a

Re: [kubernetes-users] Shared Kubernetes Control Plane

Thank you, Tim, for the reply. I sure better understand your logic. Given I am new to Kubernetes, would you please shed some lights on where the technical challenges are for such architecture (shared control plane)? Looking at the overall Kubernetes design, it seems like that the worker nodes j

Re: [kubernetes-users] Shared Kubernetes Control Plane

I think math there is a bit suspect. Yes, you'd have fewer master *nodes* in one big cluster, but you should need approximately the same amount of cores/RAM--those scale more or less with the number of nodes/pods in your cluster. (This doesn't hold completely if your individual clusters are so smal

Re: [kubernetes-users] Shared Kubernetes Control Plane

On Wed, Sep 28, 2016 at 11:54 AM, Mike wrote: > Hi Tim, > > Thank you for the answer. The goal is to share the control plane among ,say, > 100 smaller clusters (only worker nodes in each cluster) which will save you > something like 100*3=300 control plane nodes so it seems to be significant > unl

Re: [kubernetes-users] Shared Kubernetes Control Plane

Hi Tim, Thank you for the answer. The goal is to share the control plane among ,say, 100 smaller clusters (only worker nodes in each cluster) which will save you something like 100*3=300 control plane nodes so it seems to be significant unless I am missing something when you refer to it as not

Re: [kubernetes-users] Shared Kubernetes Control Plane

This is another variant of multi-tenancy, which is not a first-class supported thing in Kubernetes yet. You're actually making it vastly more complicated by describing it as multiple cloud accounts. That implies different pools of VMs, which implies different clusters. I think federation is what

Re: [kubernetes-users] Shared Kubernetes Control Plane

Essentially the workers in one group should have no way of knowing anything or affecting anything in other groups. Actually, each group can also be in a separate cloud account and so this isolation already exists at the cloud level. It is just sharing the control plane (which also lives in a di

Re: [kubernetes-users] Shared Kubernetes Control Plane

What kind of isolation are you talking about? Network isolation? Do you need network encapsulation? Resource (noisy neighbor) isolation? Is nothing shared between the groups? e.g. volumes etc.? Do pods within groups need to be broken up into namespaces? or can namespaces be the way you define group

[kubernetes-users] Shared Kubernetes Control Plane

I am new to Kubernetes and I have a question regarding the possibility of sharing the control plane of a single Kubernetes cluster among a whole bunch of workers that are - grouped into 100% isolated groups - each workers group does not need to access any pod in any other group and ac