No Route to host, no DNS taken from WebUI

[nicedevil]

Hey Guys,

i have setup a new LXC with Ubuntu 22.04.
My host is setup to use my adguard as its DNS server. All other LXCs should use exactly the hosts settings for DNS.
you guessed it, the /etc/resolv.conf is showing the gateways IP instead for the nameserver.

Tested this on 3 other test LXC and the same behavior. My older LXCs got the right settings there.

Now I have an other issue. The Ubuntu LXC from above is not able to access anything in my local network, also not in the same subnet. The firewall from Proxmox isn’t activated. I couldn’t access an other LXC via SSH or even I can’t ping it.

no route to host…

was there anything changed in Proxmox 8.0.4?

Thank you in advance for helping me out.

 

[t.lamprecht]

My host is setup to use my adguard as its DNS server. All other LXCs should use exactly the hosts settings for DNS.
you guessed it, the /etc/resolv.conf is showing the gateways IP instead for the nameserver.

Can you please post the Proxmox VE hosts (maybe slightly censored, if there's identifiable info in there) /etc/resolv.conf and a config of an affected container here (pct config VMID)?

Now I have an other issue. The Ubuntu LXC from above is not able to access anything in my local network, also not in the same subnet. The firewall from Proxmox isn’t activated. I couldn’t access an other LXC via SSH or even I can’t ping it.

The Proxmox VE's /etc/network/interfaces (I'd recommend censoring any public IP, e.g., replace the middle two octects with a letter or so) would be great to have here to.

was there anything changed in Proxmox 8.0.4?

Not really, and I don't know of any other such recent reports. Maybe there's a (rogue) DHCP server running in the network, or the newer CTs showing the issue have another network bridge selected (just guesstimating here).

 

[nicedevil]

Can you please post the Proxmox VE hosts (maybe slightly censored, if there's identifiable info in there) /etc/resolv.conf and a config of an affected container here (pct config VMID)?


The Proxmox VE's /etc/network/interfaces (I'd recommend censoring any public IP, e.g., replace the middle two octects with a letter or so) would be great to have here to.


Not really, and I don't know of any other such recent reports. Maybe there's a (rogue) DHCP server running in the network, or the newer CTs showing the issue have another network bridge selected (just guesstimating here).

First thank you for taking a bit of your time for my issue.

Here is what you wanted:

/etc/resolv.conf

root@gateway:~# cat /etc/resolv.conf
search mydomain.local
nameserver 10.0.42.2
nameserver 10.0.42.3

pct config 47007

root@gateway:~# pct config 47007
arch: amd64
cores: 4
hostname: testLXC
memory: 2048
net0: name=eth0,bridge=vmbr0,hwaddr=4A:97:41:7A:5D:F2,ip=dhcp,tag=47,type=veth
ostype: ubuntu
rootfs: local-lvm:vm-47007-disk-0,size=10G
swap: 512
tags: 47
unprivileged: 1

/etc/network/interfaces

root@gateway:~# cat /etc/network/interfaces
# network interface settings; autogenerated
# Please do NOT modify this file directly, unless you know what
# you're doing.
#
# If you want to manage parts of the network configuration manually,
# please utilize the 'source' or 'source-directory' directives to do
# so.
# PVE will preserve these directives, but will NOT read its network
# configuration from sourced files, so do not attempt to move any of
# the PVE managed interfaces into external files!

auto lo
iface lo inet loopback

iface enp3s0 inet manual

iface enp5s0 inet manual

iface enp6s0 inet manual

iface enp7s0 inet manual

iface enp8s0 inet manual

iface enp9s0 inet manual

auto vmbr0
iface vmbr0 inet static
        address 10.0.75.2/24
        gateway 10.0.75.1
        bridge-ports enp3s0
        bridge-stp off
        bridge-fd 0

There is no rogue DHCP in my network. I have a Unifi Dreammachine on my lab. The LXC got a different IP on the first boot, then I setup a reservation my UDM for this machine, and the machine took that new IP after a reboot.

EDIT: What I was able to test is a friends Proxmox environment. He also got a UDM and got different VLANs and AdGuard as its DNS. On his site the testLXC got the DNS server that is setup on the host. Maybe there is a limitation on the number of LXCs? I have 22 LXCs and 3 VMs running on the first host.

EDIT2: 2nd difference to my friends environment is, that I have a cluster (without HA) and he doesn't.

 

[nicedevil]

Ok first thing is solved now... just found out that the host I was trying to ping got only a subnet /29 instead of the correct /28....
The problem with the DNS still exists, maybe you can help with that?