Cant keep bridge up - goes DOWN and wont go UP

[magnum]

Hello Forum,

i have some weird issue. I have one host and two VMs. All are connected with two bridges (see the interfaces).
Now the link works for some time and then goes down, i have to reboot the node to get it up again. After some time it goes down again.

vmbr0 is for LAN and vmbr1 is for WAN over NAT.

ifdown and ifup:
https://pastebin.com/H7zQri7x

ip l:
https://pastebin.com/TBFGSnUW

syslog with grep:
https://pastebin.com/NmJ6fHVg


pve-manager/7.0-11/63d82f4e (running kernel: 5.11.22-4-pve)

 

[magnum]

Anyone?

 

[cppmonkey]

I had this, found that upgrading from PVE 6 to 7 installed NetworkManager and some of the NICs were configured within nmcli.
Deleted the configuration from NM and restarted networking and everything was good.

I was pulling my hair out at why 1 LACP would come up but not the other.

Now I can't exceed 10G (9.36G) of throughput because of the veth driver!

 

[cppmonkey]

Looks like PVE7 is starting to use NetworkManager too.
I moved away from ifupdown years ago. Was weird inheriting some PVE6 servers and having to use it all over again

 

[spirit]

Looks like PVE7 is starting to use NetworkManager too.
I moved away from ifupdown years ago. Was weird inheriting some PVE6 servers and having to use it all over again

? proxmox don't use networkmanager. it's using ifupdown1 or ifupdown2.

 

[spirit]

Hello Forum,

i have some weird issue. I have one host and two VMs. All are connected with two bridges (see the interfaces).
Now the link works for some time and then goes down, i have to reboot the node to get it up again. After some time it goes down again.

vmbr0 is for LAN and vmbr1 is for WAN over NAT.

ifdown and ifup:
https://pastebin.com/H7zQri7x

ip l:
https://pastebin.com/TBFGSnUW

syslog with grep:
https://pastebin.com/NmJ6fHVg


pve-manager/7.0-11/63d82f4e (running kernel: 5.11.22-4-pve)

a bridge will go down if no interfaces is plugged in the bridge. (it should be the case when all your vms are stopped).

maybe can you try to put

        post-up   iptables -t nat -A POSTROUTING -s '10.10.10.0/24' -o eno1 -j MASQUERADE
        post-down iptables -t nat -D POSTROUTING -s '10.10.10.0/24' -o eno1 -j MASQUERADE

under eno1 interface to be sure.

 

[cppmonkey]

Checking nmcli, it shows the interfaces for my CTs as `managed`
Everything else is `external/unmanaged`

 

[spirit]

Checking nmcli, it shows the interfaces for my CTs as `managed`
Everything else is `external/unmanaged`

nmcli / networkmanager is not installed in a normal proxmox installation. proxmox don't install it.

 

[magnum]

I have not installed nmcli.I'm using ifupdown2 from repo:
apt show ifupdown2
Package: ifupdown2
Version: 3.1.0-1+pmx3

@spirit Im not really sure what the difference would be. Perhaps its due to my iptables havinge redundant information? should i clear iptables?

iptables -t nat -L Chain PREROUTING (policy ACCEPT) target prot opt source destination DNAT tcp -- anywhere anywhere multiport dports !ssh,8006 to:10.0.0.2 DNAT udp -- anywhere anywhere to:10.0.0.2 DNAT tcp -- anywhere anywhere multiport dports !ssh,8006 to:10.0.0.2 DNAT udp -- anywhere anywhere to:10.0.0.2 DNAT tcp -- anywhere anywhere multiport dports !ssh,8006 to:10.0.0.2 DNAT udp -- anywhere anywhere to:10.0.0.2 Chain INPUT (policy ACCEPT) target prot opt source destination Chain OUTPUT (policy ACCEPT) target prot opt source destination Chain POSTROUTING (policy ACCEPT) target prot opt source destination MASQUERADE all -- 10.0.0.0/24 anywhere


EDIT: Flushed, but still linkdown. Ive added some bridges in interfaces: https://pastebin.com/9pLz44RH
Goal: https://schroederdennis.de/allgemei...ver-mit-nur-1-public-ip-addresse-pfsense-nat/

 

[magnum]

brctl cant delete vmbr2 either if its up or down..

 

[magnum]

If i do :
service networking restart
i'll lose all living connections to tapXXX and therefore they go down. Restarting the vm seems to bring them back

 

[spirit]

If i do :
service networking restart
i'll lose all living connections to tapXXX and therefore they go down. Restarting the vm seems to bring them back

This is expected, don't do networking restart on proxmox, only reload.
(vm tap interfaces are not managed in /etc/network/interfaces, so they are lost if you restart network)