Ok until late last year i'm a complete Linux newbie so do excuse me if my questions seem rudimentary and somewhat bizarre/silly as the 1st thing after i tried to educate myself on the various linux distros and features is to try to understand Virtualisation. I've only played with QEMU/Virtualbox inside a linux OS but now i am seriously considering a Type 1 Hypervisor and installing linux distro as my 24/7 OS and Windows 10 as a back up VM in case i absolutely have to run windows only software? Now i am starting to learn about passing thru GPUs and other stuff but i am still confused about a HV i am keen on namely Proxmox and its "LXC Containers" features? My basic questions are

#1 These templates are "mini OSes" that rely on Proxmox linux kernel to run apps?

#2 How much resources does a user know to allocate to template/containers?

#3 Can i have one machine with Proxmox + LXC template installed serve multiple clients at the same time? Do they consume the same or less resources compared to a VM?

#4 Can a machine run VMs *and* LXC Containers at the same time serving local and remote clients?

That's a good one-line characterization. They contain a fairly complete operating system except for the kernel.
This sentence doesn't make sense to me. Can you rephrase it?
Yes you can, and they consume considerably fewer resources than VMs, factor ten perhaps (depending on the workload obviously).
I don't know how Proxmox handles this, but in principle yes. I know that OpenStack only permits one type of hypervisor per physical server.

1 members found this post helpful.

Thanks a lot that was useful but i've been snooping around the net but best i explain what i'm trying to achieve as being both a linux *and* VM/Containers newbie the technical terms i use may not be accurate or describe exactly my intentions

Basically imagine a (desired) scenario

10-20 fat clients "worker nodes"
1 centralised VM/Compute Cloud server "master node"


The master node stores all critical files that worker nodes access. Ideally all worker nodes run off an OS that only the master node can update, patch, etc Can worker nodes do "superficial" things like change desktop wallpaper, how icons are arranged on what they see on desktop when logging into the OS on master node?

Also since worker nodes are fat clients from my snooping around these few days, seems the master node need not be a huge order more powerful as most of the processing is done by the fat clients correct? Even smartphones/tablets are powerful enough to run simple data entry/office apps these days and some worker nodes are only doing this. These proxmox templates from what i understand can run on both x86 and ARM?

So basically my intention is:

Master node runs a VM which could be Ubuntu, OpenSuse or Fedora best if it can give access to fat clients mostly on x86 arch but some on ARM?

Fat clients have OS of their own like Windows, Mac, Android, etc Can i install Proxmox on USB drive and preconfigure them so that worker nodes receive these and all they have to do is boot off them on their machines (fat client) , key in the user/login master has created for them and do the work? How does a VM on cloud factor in different arch or hardware drivers/devices on clients? For example 1-2 worker nodes may specialise in CAD/CAM and have a Radeon Pro/Quadro discrete GPU - does the Master Node VM have special worker node folders to allow for worker drivers to be installed or are they installed on worker node drive (the usb?) Or does one have to set up VM in a master node with all combo of hardware/drivers downloaded/configured?

Lastly only 2-3 worker nodes share same network - the rest are peeps working from their homes or on the go so the VM is set up to be accessed from LAN same network or from the net?

Hope my explanation above is clearer - could be that the solution does not even involve Proxmox but seeing as this is a small work group going free/open source sounds like an excellent idea ^^