I am trying to install KVM virtual machines using Yast and am having difficulties. I have experience using VirtualBox but I’m new to using KVM and Yast Virtual Machines. My problem is that I have tried three times to install a vm but the installations failed for various reasons. (Probably because I also selected using an Ethernet Bridge mode which is not recommended). I could not find examples for how to delete these failed installations. I was able to find virsh commands to delete domains, and I did that for two of the three vm attempts. I think the third attempt might actually have been a copy of one of the first two attempts. And I’m uncertain if I’ve cleaned up all the files and file systems that were created.
Other than reinstalling my Leap 15.1. Are there file locations I could check and delete files to clean up these failed VM attempts? I have not found any reference for where Yast puts these virtual machine files. (Unlike virtualBox which located them in my home folder).
It also might help to know what program to run to start and stop and display these virtual machines. It was easy for me to do so using virtualBox, however I have been unable to find what to use in Leap 15.1 to actually use the vms created under Yast. Will some Yast app become available after a vm is successfully installed and running? Using virsh I did see that one of my vms was running but I didn’t know how to then actually use it.
I assume that since KVM is integrated into Leap, and uses virtualization onboard the microprocessor thatVir it will run more efficiently and faster than using VirtualBox.
Hi
Install and use virt-manager to setup your virtual machines, I would recommend a separate partition for the images (xfs format) as in /var/lib/libvirt/images
The install wizard should help configure/tweak the VM based on the distribution your installing.
In virt-manager you will also be able to delete the images and clean up the supporting files.
I am seeing the definitions installed under “/etc/libvirt” and particularly at “/etc/libvirt/qemu”.
I am seeing the virtual disks installed at “/var/lib/libvirt/images”.
I am seeing the UEFI NVRAM installed at “/var/lib/libvirt/qemu/nvram”.
If you manually remove files, it might be a good idea to restart libvirt afterwards. I normally use “virt-manager” to remove unwanted VMs.
Thank you all; I was able to use virt-manager to delete two vms, and the second vm did have a clone that needed to be deleted before it was deleted. Now I just need to figure out why my next trys are failing… But I 'll try virt-manager to install it.
Currently the directories mentioned no longer appear to have any files associated with my mistakes in them. And virt-manager no longer shows any machines available or running.
Also I may have mistakenly stated that I’m running Leap 15.1 while I’m actually on Leap 15.0.
I am sure I’ll have more questions. For instance do we use virt-manager to start and stop virtual machines, but then require virt-viewer to view the console or display for the virtual machine that we just started? Also then do we need to use virt-viewer to run the installation scripts?
I am also confused by the statement that virt-manager includes access to a graphical console of the VM Guest using VNC. Does this mean that we use virt-viewer if we want to see the graphical console of a VM running on another machine while we use virt-manager to access the console if we started the machine locally?
A link to examples on setting up and running VMs using virt-manager would be awesome. I’ll search for them next. Searching for setting up VMs with Leap 15.0 came back with KVM and Yast.
Normally, you should only use one or two libvirt apps
vm manager should be your primary tool for listing, managing and start/stopping your Guests.
When you create a new vm, you can launch virt install from your HostOS app menu or from vm manager.
Anything else like virt viewer can be, but is not normally invoked on its own.
So, for instance when you start a Guest in vm manager, virt viewer should automatically be invoked, ordinarily using the VNC protocol to display your Guest.
The following vm manager webpage should clarify what vm manager uses to perform various tasks… https://virt-manager.org/
As for your reasons for using KVM…
KVM is an “enterprise” type solution which means ti’s management tools and approach is intended to support Production use with VMs running not necessarily interactively (Logged on Users). Although you can deploy many workstation type Guests, you’ll probably find Guests running as Servers instead, providing network services and application functionality to a client machines.
Virtualbox on the other hand is not designed for Production servers running unattended, in fact it’s installed as a User application not intended to be left running when the User logs off. Tools might be designed to be more User-friendly keeping in mind that Virtualbox Users might be less experienced.
As for performance if that is what you might mean by “efficiency,” there is likely little difference. Both KVM and Virtualbox use hardware CPU extensions mostly the same way so there is likely hardly any difference in performance.
Bottom line…
Aside from whether you just want to learn KVM,
Whether anyone would choose KVM or Virtualbox should depend on what they intend to do with their virtual machines, whether you intend to deploy Servers that provide services to your network or whether you want to run something private for your own personal use.