Connection is lost intermittently and IPv4 address disappears for both supervisor and VMs

[linuswaerner]

Hi, I've scanned the forums and read hundreds of threads here and on other forums as well - but nothing seems to help.

I'm running a Proxmox server and it's connected to an Unifi UDM with three networks: 10.0.0.0/24, 192.168.0.0/24 and 172.16.0.0/24.
The VLANS are tagged with 192 and 172 (which corresponds to the first block of it).

Here's the interfaces file:

auto lo
iface lo inet loopback

iface enp2s0 inet manual

auto vmbr0
iface vmbr0 inet static
        address 10.0.0.99/24
        gateway 10.0.0.1
        bridge-ports enp2s0
        bridge-stp off
        bridge-fd 0
        bridge-vlan-aware yes


auto vmbr0.172
iface vmbr0.172 inet static
        address 172.16.0.0/24


auto vmbr0.192
iface vmbr0.192 inet static
        address 192.168.0.0/24

The problem is that all IPv4 connections dies intermittent. It can run for 100 days, but after a reboot or anything that affects its state, it won't get back online. I restart the router, the server itself, the VMs - nothing. Then, all of a sudden it can just get back to life again.
I haven't been able to find anything out of the ordinary in any logs.
Earlier tonight it started working, but all of a sudden it died again once I started a VM.

When it doesn't work, this is the output from ip address

enp2s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        ether 54:04:a6:12:99:2a  txqueuelen 1000  (Ethernet)
        RX packets 42000  bytes 14165689 (13.5 MiB)
        RX errors 0  dropped 116  overruns 0  frame 0
        TX packets 37706  bytes 17471881 (16.6 MiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
       
vmbr0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet6 fe80::5604:a6ff:fe12:992a  prefixlen 64  scopeid 0x20<link>
        ether 54:04:a6:12:99:2a  txqueuelen 1000  (Ethernet)
        RX packets 12997  bytes 2839025 (2.7 MiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 10738  bytes 4521362 (4.3 MiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

The key difference is that when it's operational, I also get this little badboy:
inet 10.0.0.99 netmask 255.255.255.0 broadcast 10.0.0.255

The issue is the same for both the supervisor and it's hosted VMs. All VMs have reserved IP addresses set in the UDM as well.

Is there something wrong with my interfaces file? What else can I look for?