With free esxi over, not shocking bit sad, I am now about to move away from a virtualisation platform i’ve used for a quarter of a century.
Never having really tried the alternatives, is there anything that looks and feels like esxi out there?
I don’t have anything exceptional I host, I don’t need production quality for myself but in all seriousness what we run at home end up at work at some point so there’s that aspect too.
Thanks for your input!
- KVM/QEMU/Libvirt/virt-manager on a Debian 12 for minimal installation that allows you to choose backup tools and the like on your own.
- Proxmox for a mature KVM-based virtualizer with built in tools for backups, clustering, etcetera. Also supports LXC. https://github.com/proxmox
- Incus for LXC/KVM virtualization - younger solution than Proxmox and more focused on LXC. https://github.com/lxc/incus
/thread
This is my go-to setup.
I try to stick with libvirt/
virsh
when I don’t need any graphical interface (integrates beautifully with ansible [1]), or when I don’t need clustering/HA (libvirt does support “clustering” at least in some capability, you can live migrate VMs between hosts, manage remote hypervisors from virsh/virt-manager, etc). On development/lab desktops I bolt virt-manager on top so I have the exact same setup as my production setup, with a nice added GUI. I heard that cockpit could be used as a web interface but have never tried it.Proxmox on more complex setups (I try to manage it using ansible/the API as much as possible, but the web UI is a nice touch for one-shot operations).
Re incus: I don’t know for sure yet. I have an old LXD setup at work that I’d like to migrate to something else, but I figured that since both libvirt and proxmox support management of LXC containers, I might as well consolidate and use one of these instead.
I use cockpit and my phone to start my virtual fedora, which has pcie passthrough on gpu and a usb controller.
Desktop:
Mobile:
We use cockpit at work. It’s OK, but it definitely feels limited compared to Proxmox or Xen Orchestra.
Red Hat’s focus is really on Openstack, but that’s more of a cloud virtualization platform, so not all that well suited for home use. It’s a shame because I really like Cockpit as a platform. It just needs a little love in terms of things like the graphical console and editing virtual machine resources.
Ooh, didn’t know libvirt supported clusters and live migrations…
I’ve just setup Proxmox, but as it’s Debian based and I run Arch everywhere else, then maybe I could try that… thanks!
In my experience and for my mostly basic needs, major differences between libvirt and proxmox:
- The “clustering” in libvirt is very limited (no HA, automatic fencing, ceph inegration, etc. at least out-of-the box), I basically use it to 1. admin multiple libvirt hypervisors from a single libvirt/virt-manager instance 2. migrate VMs between instances (they need to be using shared storage for disks, etc), but it covers 90% of my use cases.
- On proxmox hosts I let proxmox manage the firewall, on libvirt hosts I manage it through firewalld like any other server (+ libvirt/qemu hooks for port forwarding).
- On proxmox I use the built-in template feature to provision new VMs from a template, on libvirt I do a mix of
virt-clone
andvirt-sysprep
. - On libvirt I use
virt-install
and a Debian preseed.cfg to provision new templates, on proxmox I do it… well… manually. But both support cloud-init based provisioning so I might standardize to that in the future (and ditch templates)
My understanding is that for proper cluster management you slap Pacemaker on there.
they need to be using shared storage for disks
You can perform a live migration without shared storage with libvirt
I should RTFM again… https://manpages.debian.org/bookworm/libvirt-clients/virsh.1.en.html has options for
virsh migrate
such as--copy-storage-all
… Not sure how it would work for actual live migrations but I will definitely check it out. Thanks for the hintPretty darn well. I actually needed to do some maintenance on the server earlier today so I just migrated all of the VMs over to my desktop, did the server maintenance, and then moved the VMs back over to the server, all while live and functioning. Running ping in the background looks like it missed a handful of pings as the switches figured their life out and then was right back where they were; not even long enough for uptime-kuma to notice.
I didn’t know libvirt supported HA
Did you read? I specifically said it didn’t, at least not out-of-the-box.
libvirt does support “clustering” at least in some capability, you can live migrate VMs between hosts
clustering != HA
The “clustering” in libvirt is limited to remote controlling multiple nodes, and migrating hosts between them. To get the High Availability part you need to set it up through other means, e.g. pacemaker and a bunch of scripts.
Also VirtualBox.
They’re obviously looking for a type 1 hypervisor like Esxi. A type 2 hypervisor like virtualbox does not fit the bill.
What is the difference between type 1 & 2 please ?
Type 1 runs on bare metal. You install it directly onto server hardware. Type 2 is an application (not an OS) lives inside an OS, regardless of whether that OS is a guest or a host, the hypervisor is a guest of that platform, and the VMs inside it are guests of that hypervisor.
Thank you
The previous comment is an excellent summary. It is worth noting that there are some type 1 hypervisors that can look like type 2s. Specifically, KVM in Linux (which sometimes gets referred to as Virt-manager, Virtual Machine Manager, or VMM, after the program typically used to manage it) and Hyper-V in Windows.
These get mistaken for type 2 hypervisors because they run inside of your normal OS, rather than being a dedicated platform that you install in place of it. But the key here is that the hypervisor itself (that is, the software that actually runs the VM) is directly integrated into the underlying operating system. You were installing a hypervisor OS the whole time, you just didn’t realise it.
The reason this matters is that type 1 hypervisors can operate at the kernel level, meaning they can directly manage resources like your memory, CPU and graphics. Type 2 hypervisors have to queue with all the other pleb software to request access to these resources from the OS. This means that type 1 hypervisors will generally offer better performance.
With hypervisor platforms like Proxmox, Esxi, Hyper-V server core, or XCP-NG, what you get is a type 1 hypervisor with an absolutely minimal OS built around it. Basically, just enough software to the job of running VMs, and nothing else. Like a drag racer.
VB is awful.
And I use it every day.
It’s like a first-try at a hypervisor. Terrible UI, with machine config scattered around. Some stuff can only be done on the command line after you search the web for how to do it (like basic stuff, say run headless by default). Enigmatic error messages.
This is what I would recommend too - QEMU + libvirt with Sanoid for automatic snapshot management. Incus is also a solid option too
Proxmox works well for me
proxmox
This is the way
If you’re running mostly Linux vms proxmix us really good. It’s based on kvm and has a really nice feature set.
Windows guests also run fine on KVM, use the Virtio drivers from Fedora project.
I’ve used Hyper-V and in fact moved away from ESXi long ago. VMWare had amazing features but we could not justify the ever-increasing costs. Hyper-V can do just about anything VMWare can do if you know Powershell.
Another vote for Hyper-V. Moved to it from ESXi at home because I had to manage a LOT of Hyper-V hosted machines at work, so I figured I’d may as well get as much exposure to it as I could. Works fine for what I need.
Seconded for Hyper-V, and MUCH easier to patch the free edition than ESXi.
I use it with WAC on my home server and it’s good enough for anything I need to do. Easy to create VMs using that UI, PS not even needed.
I’m pretty happy with XCP-ng with their XenOrchestra management interface. XenOrchestra has a free and enterprise version, but you can also compile it from source to get all the enterprise features. I’d recommend this script: https://github.com/ronivay/XenOrchestraInstallerUpdater
I’d say it’s a slightly more advanced ESXi with vCenter and less confusing UI than Proxmox.
I second this and would like to add in another resource for XCP-ng : Lawrence Systems on YouTube.
Qemu/virt manager. I’ve been using it and it’s so fast. I still need to get the clipboard sharing working but as of right now it’s the best hypervisor I’ve ever used.
I love it. Virtmanager connecting over ssh is so smooth.
Acronyms, initialisms, abbreviations, contractions, and other phrases which expand to something larger, that I’ve seen in this thread:
Fewer Letters More Letters DNS Domain Name Service/System ESXi VMWare virtual machine hypervisor HA Home Assistant automation software ~ High Availability LTS Long Term Support software version LXC Linux Containers ZFS Solaris/Linux filesystem focusing on data integrity k8s Kubernetes container management package
[Thread #540 for this sub, first seen 24th Feb 2024, 11:35] [FAQ] [Full list] [Contact] [Source code]
I do do not not know— however, that logo is amazing
EDIT: Found it — https://sega-ai.neocities.org/
If you are dipping toes into containers with kvm and proxmox already, then perhaps you could jump into the deep end and look at kubernetes (k8s).
Even though you say you don’t need production quality. It actually does a lot for you and you just need to learn a single API framework which has really great documentation.
Personally, if I am choosing a new service to host. One of my first metrics in that decision is how well is it documented.
You could also go the simple route and use docker to make containers. However making your own containers is optional as most services have pre built ones that you can use.
You could even use auto scaling to run your cluster with just 1 node if you don’t need it to be highly available with a lot of 9s in uptime.
The trickiest thing with K8s is the networking, certs and DNS but there are services you can host to take care of that for you. I use istio for networking, cert-manager for certs and external-dns for DNS.
I would recommend trying out k8s first on a cloud provider like digital ocean or linode. Managing your own k8s control plane on bare metal has its own complications.
There are also full-suites like rancher which will abstract away a lot of the complexity
K8s is great, but you’re chaning the subject and not answering OPs question. Containers =/= VMs.
You are right. But proxmox and many of the other suggestions aren’t vms either.
I actually moved everything to docker containers at home… Not an apples to apples, but I don’t need so many full OSs it turns out.
At work we have a mix of things running right now to see. I don’t think we’ll land on ovirt or openstack. It seems like we’ll bite the cost bullet and move all the important services to amazon.
OOTL and someone who only uses a vm once every several years for shits & grins: What happened to vmware?
As part of the transition of perpetual licensing to new subscription offerings, the VMware vSphere Hypervisor (Free Edition) has been marked as EOGA (End of General Availability). At this time, there is not an equivalent replacement product available.
For further details regarding the affected products and this change, we encourage you to review the following blog post: https://blogs.vmware.com/cloud-foundation/2024/01/22/vmware-end-of-availability-of-perpetual-licensing-and-saas-services/
Whelp…boo-urns. :(
Minikube and try to get everything on Kubernetes?
Where does running VMs compare in any way to what Kubernetes does?
Depends on what you want to self host? Could be worth it to see if what you self host can be deployed as containers instead
Kubernetes yes, but minikube is kinda meh as a way to install it outside of development environments.
There’s so many better manageable ways like RKE/Rancher (which gives you the possibility to go k3s),Kubespray or even kubeadm.
All of those will result in a cluster that’s more suitable for running actual workloads.
I wouldn’t recommend going K8S only in a homelab. Too much effort and some things don’t fit well (Home Assistant, Gaming VM?)
It’s just a suggestion. Many would probably find that the workload they host is available on containers. I run a Kubernetes cluster on bare metal at home. There’s also nothing stopping you from creating VMs that you can ssh to with KubeVirt
Don’t get me wrong, I like the idea. But personally it never worked out for me.
I know everyone says to use Proxmox, but it’s worth considering xcp-ng as well.
In my “testing” at work and private, PVE is miles ahead of xcp-ng n terms of performance. Sure, xcp-ng does it’s thing very stable, but everything else…proxmox is faster
I agree that Proxmox VE is better; I’m just saying that people should compare multiple options and pick the one they like the best.
I’m using Unraid on my home server because it can run Docker containers in addition to KVM and LXC (via a plugin).
For home have a crack at KVM with front ends like proxmox or canonical lxd manager.
In an enterprise environment take a look at Hyper-V or if you think you need hyper converged look at Nutanix.