Bonjour aux bienfaiteurs

J’ai une VM Ubuntu que refuse obstinément de communiquer avec le réseau depuis le clash avec le « cache » qui était saturé.

Pour le cache cela est maintenant solutionné grâce à l’accompagnement de JorgeB encore merci

 

Tous les détails sur ce post : 

 

 

Cette VM  "Ubuntu serveur web" fonctionnait parfaitement avant l’incident avec le cache.

Impossible de monter une carte réseau

 

 

Dans les log du gestionnaire de VM j'ai ce type d'erreur 

 

 

Il y a aussi des VM Windows dont je ne recontre aucun incident.

 

D'avance merci pour votre coups de main

 

EricW

 

 

 

 

tower-diagnostics-20220601-2005.zip

Hello, 
Help me!

On 6/1/2022 at 9:11 PM, EricW said:

Dans les log du gestionnaire de VM j'ai ce type d'erreur 

Hi,

that are related to qemu guest agent, which is probably not installed in the guest; the qemu guest agent is not mandatory, its functions are to manage some services of the virtual machine, like shutdown, reboot, etc without relaying on acpi commands.

 

As far as the network is concerned there's nothing wrong in the network settings, the virtual machine is using a virtio-net virtual adapter, br0 bridged to eth0, network 192.168.0.x/255.255.255.0.

 

I would start to:

1. see in the guest the available networks/ip addresses with ifconfig or ip addr commands

2. maybe your guest lacks virtio drivers, install virtio drivers or change in the xml the network type from virtio-net to something standard, like e1000-82545em.

    <interface type='bridge'>
      <mac address='52:54:00:1e:82:b3'/>
      <source bridge='br0'/>
      <model type='e1000-82545em'/>
      <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x0'/>
    </interface>

 

Edited June 3, 20224 yr by ghost82

Hello ghost82,

 

J'ai modifié le XML, et désolé, cela ne fonctionne toujours pas.

 

J'ai fait le test d'arrêter et de redemarrer le service "Gestionnaire des VM" => j'ai le message suivant 

 

J'ai redémarrer le serveur Unraid.

Le service a bien redémarré, mais il y a toujours les mêmes erreurs dans le log LIBVIRT, la VM a bien redémarré mais de carte reseau n'est pas remonté...

 

EricW

 

 

Edited June 4, 20224 yr by EricW

Run:

lspci

in a terminal of the Ubuntu serveur web vm.

 

Run:

ip addr

in a terminal of the Ubuntu serveur web vm.

 

Paste here the outputs.

 

Please write in english, otherwise this post will be moved somewhere else.

Hello Ghost82

 

ok, as you can see the ethernet controller is there, at 02:01.0, named ens1 (mac: 52:54:00:1e:82:b3).

So, for what it concerns the vm settings there's nothing wrong, and the "issue" is in ubuntu itself

The interface is down, how are you managing connections? with networkmanager or with networkd?

Try to put the interface up:

ip link set dev ens1 up

then rerun the ip addr comamnd and see if it's up and if it gets an ip.

Voici les différentes étapes que j'ai réalisé :

 

(malgrés le DHCP, la carte ne prend pas de IP, j'ai forcé l'IP et pour autant le reseau n'est tourjours pas visible)

 

 

 

What is 192.168.0.250?

br0 interface on unraid has ip 192.168.0.203.

If 192.168.0.250 is online it could be that ping requests are blocked by a firewall, for example if I remember well pinging a windows 10 machine doesn't work by default because of ICMP settings.

This is an issue with your network or your ubuntu installation and not related to vms.

The fact that your router has settings for dhcp doesn't mean that the vm will get an ip if it's misconfigured.

 

Again, if you want to continue commenting, write in english, use a translator if you aren't able to write in english.

Hello Ghost82

The 192.168.0.250 is the PfSense firewall.

For security I checked if the Unraid server pings the firewall, the ICMP rule is not in place.

 

For security I checked if the Unraid server pings the firewall, the ICMP rule is not in place.

The Windows VM pings through the lan the firewall. I installed a new Ubuntu VM with the following interface settings:

 

    <interface type='bridge'>
      <mac address='52:54:00:34:25:05'/>
      <source bridge='br0'/>
      <model type='virtio-net'/>
      <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x0'/>
    </interface>

 

At the first start of the VM, the network is correctly mounted, and the VM was visible on the LAN and it communicated on the WEB.

After modifying the interface to <model type='e1000-82545em'/> model, impossible on this VM to reassemble the network, and even after having reinstalled on <model type='virtio-net'/>

 

I think that the test carried out rules out the network problem?

Really no idea....I'm sorry.

e1000* adapters are the oldest ones and they should be the most compatible, in my linux boxes I have no issues with these types of adapters.

I would check the firewall configuration, or maybe there are issues in the os or issues with network configuration inside the os.

As far as the vm, the adapter is recognized and plugged, why it's not up on boot...I don't know..

The network config hasn't changed for months.
On Unraid is it not possible to uninstall the VM service and reinstall it?
Before the problem with the pool, I had no problem, I installed Ubuntu VM several times without having any problem.