I’m following the “Adding Virtualization Support” Fedora Server doc and can’t make sense of some parts of the libvirt default network configuration file.
I have a home lab using Opnsense on it’s own physical machine (nothing else on it). It is for my local DNS, DHCP and firewall services.
I have one network provisioned with all of my computers on it (I’m not doing separate virtual networks for IoT right now - maybe later).
I am trying to setup one of my physical computers (Fedora Server 42) to be used as a host for virtual machines (KVM/QEMU/Libvirt).
I’m struggling with knowing the correct entries for some tags in the libvirt default network configuration file (virsh net-edit default).
The example file that I’m working with has the usual tagAlong with the other usual tags in that that file.
Here are tags that are confusing me:
In the “network” section:
“domain name=‘example.lan’ localOnly=‘yes’” Question 1: Do I make up my own fictitious domain name that is different from the domain name that I’m using in Opnsense?
In the “dns” section:
“forwarder domain=‘example.lan’”
I would assume that I’d use whatever goes in tag above.
“host ip='192.168.122.1”
My existing network is using 192.168.1.xxx per the Opnsense DHCP service. Question 2:
Is it okay to use the example ip address of 192.168.122.1 that the example gives me?
“hostname"host"hostname” Question 3: Is this the hostname of the physical server or some made up name that won’t be used by anything else?
“hostname"host.example.lan"hostname”
I can get this once I get domain name and hostname figured out.
I can also figure out the ip address, netmask tag and dhcp range start/end once I get the tag figured out.
I’ve googled, consulted AI models and still struggling with this part.
[Edit:]
After posting my original post: I went to my server’s cockpit and looked at the Networking tab. I show an already existing interface named “br0” with an ip address that would have been assigned by Opnsense and an interface named vibr0 with an ip address that would have been assigned by the local network that I configured when following the Fedora Server doc on how to set up virtualization.
[Original post:]
Thanks for responding.
The “nmcli connection delete” was my wired connection enp2s0.
Tee didn’t work for me. I ended up using another editor to edit bridge.conf.
The “virsh list --all” showed my running VM named VMserver1.
I used virsh shutdown to shut it down.
The “EDITOR…sed…” line didn’t work for me. However, I looked up what your command line did and went to “virsh edit VMserver1”.
There was no text with “virbr0” and only one mention of “bridge”.
I did show a virbr0 with “nmcli connection show” but it looked greyed out so, I’m guessing it was down.
Bridge showed up in:
interface type = direct
source dev=enp2s0
mode=bridge
type=virtio
I then, exited file making no changes.
Not sure what to do now since I didn’t find “virbr0” inside my VMserver1 file.
Any ideas?
Thanks.
I think my solution here is to start from scratch. I’m going to make another post with what I’m trying to do and hopefully get pointed to some places that can help me do that as well as ask more questions here. I was following this: Installing the Fedora Server Edition as a virtual machine using Cockpit
which had me doing this first: Adding Virtualization Support
However, it seems as though the “Adding Virtualization Support” process has me installing a type of network configuration that’s not optimal for having VMs on my network like regular physical servers.