libvirt users - Oct 2019 - Live migrate + change interface name

Is it possible to do a live migrate of a guest, having on the from host 
a source_device=eth2 and to host a source_dev=eth1?

On 10/17/19 4:26 PM, Marc Roos wrote:
> 
> Is it possible to do a live migrate of a guest, having on the from host
> a source_device=eth2 and to host a source_dev=eth1?
What management tool are you using that the syntax is
"source_device=eth2"?

Are you maybe just paraphrasing your config, and what you actually have 
is something like this?:

    <interface type='direct'>
      <source dev='eth2'/>
      ...
    </interface>

?

If so, the way to make this easily migratable is to create a network on 
both hosts that points to the desired physical device, e.g. on host 1:


     <network>
       <name>direct-net</name>
       <forward mode='bridge'>
         <interface dev='eth2'/>
       </forward>
     </network>

and on host2:

      (same thing, but use 'eth1')

After net-define-ing the networks, you'll need to net-autostart and 
net-start them. Then in your guest's interface config, you would use this:

      <interface type='network>
        <source network='direct-net'/>
        ...

You will then be able to migrate from one host to the other without 
needing to modify your XML during the migration.

Hmm indeed, I will have to look at using this network. Currently I am 
using virsh, and scripting the adding and removing of an interface at 
runtime with detach-interface/attach-interface.


-----Original Message-----
To: libvirt-users
Subject: Re: [libvirt-users] Live migrate + change interface name

On 10/17/19 4:26 PM, Marc Roos wrote:
> 
> Is it possible to do a live migrate of a guest, having on the from 
> host a source_device=eth2 and to host a source_dev=eth1?
What management tool are you using that the syntax is 
"source_device=eth2"?

Are you maybe just paraphrasing your config, and what you actually have 
is something like this?:

    <interface type='direct'>
      <source dev='eth2'/>
      ...
    </interface>

?

If so, the way to make this easily migratable is to create a network on 
both hosts that points to the desired physical device, e.g. on host 1:


     <network>
       <name>direct-net</name>
       <forward mode='bridge'>
         <interface dev='eth2'/>
       </forward>
     </network>

and on host2:

      (same thing, but use 'eth1')

After net-define-ing the networks, you'll need to net-autostart and 
net-start them. Then in your guest's interface config, you would use 
this:

      <interface type='network>
        <source network='direct-net'/>
        ...

You will then be able to migrate from one host to the other without 
needing to modify your XML during the migration.

I was trying to see if it is possible to automatically set the mtu size 
for the guest. I am using a tunnel with smaller mtu so when I switch 
over vm's, it would be nice if they get this tunnel mtu automatically. 

This config:
    <interface type='direct'>
      <source dev='tun1' mode='bridge'/>
      <model type='virtio'/>
      <mtu size='1462'/>
      <address type='pci' domain='0x0000' bus='0x00'
slot='0x07'
function='0x0'/>
    </interface>

Generates this error:
error: unsupported configuration: setting MTU on interface type direct 
is not supported yet

Is there another way to set this, like with this host_mtu as mentioned 
in[1], or can I forget about trying? this?

qemu-system-x86-2.0.0-1.el7.6.x86_64
qemu-kvm-1.5.3-173.el7_8.3.x86_64
libvirt-4.5.0-33.el7_8.1.x86_64

[1]
https://www.mail-archive.com/libvirt-users@redhat.com/msg10746.html

I am having a few hosts running just libvirt, among which I can do live 
migration. Is there some way or does someone have a script to balance 
guests evenly across these hosts? Based on memory usage eg.