Inaccessible, won't stay on the network, completely useless

Desktop4078

New Member
Nov 13, 2024
8
0
1
At my wits end here and extremely frustrated. It seems like, if I have a display plugged in, everything works fine. If I don't it's completely useless. My proxmox box boots up, if I'm quick I can log into the web UI, but then it will freeze and that's the end of it... completely inaccessible. No web UI, no ssh, nothing. It was running just fine for nearly 2 days with a display plugged it, then I unplugged it and now I cannot do anything anymore.

Makes zero sense. None. Not a bit. What is the problem?

IP address is assigned by router with static IP
 
There's a bios option to boot without keyboard, earlier mention as "halt on error" to disable and then you don't need a display also.
 
  • Like
Reactions: Kingneutron
IP address is assigned by router with static IP
In PVE you should assign the IP address within the node. Probably choose something out of the DHCP range of the router.

(With the monitor plugged in!) Please post output for the following:
Code:
cat /etc/network/interfaces


ip a

if I have a display plugged in, everything works fine
Can you then access via the NW through Web GUI & SSH?
 
If it works with a display, run it with a display connected. Try not to overthink it too much. Weird s--t happens in life all the time.

You mentioned NOTHING about your hardware details, how is anyone supposed to advise? Maybe your rig needs a fake-display dongle.

https://www.amazon.com/Headless-Display-Emulator-Headless-1920x1080-Generation/dp/B06XT1Z9TF?th=1
Not an option, I do not have a display that I can dedicate to it. Other Linux OSs run just fine without a display or dummy adapter connected. If a hypervisor cannot run headless then something is wrong with it. I would venture to guess that MOST proxmox deployments are headless.

It's an Intel nuc 8thgen i3
 
In PVE you should assign the IP address within the node. Probably choose something out of the DHCP range of the router.

(With the monitor plugged in!) Please post output for the following:
Code:
cat /etc/network/interfaces


ip a


Can you then access via the NW through Web GUI & SSH?

Code:
auto lo
iface lo inet loopback

iface eno1 inet manual

auto vmbr0
iface vmbr0 inet static
address 192.188.0.5/24
gateway 192.188.0.1
bridge-ports eno1
bridge-stp off
bridge-fd 0


source /etc/network/interfaces.d/*

Code:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host noprefixroute  
       valid_lft forever preferred_lft forever
2: eno1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master vmbr0 state UP group default qlen 1000
    link/ether 1c:69:7a:02:68:9e brd ff:ff:ff:ff:ff:ff
    altname enp0s31f6
3: vmbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
    link/ether 1c:69:7a:02:68:9e brd ff:ff:ff:ff:ff:ff
    inet 192.188.0.5/24 scope global vmbr0
       valid_lft forever preferred_lft forever
    inet6 fe80::1e69:7aff:fe02:689e/64 scope link  
       valid_lft forever preferred_lft forever

& yes, I can access the web UI when a screen is connected
 
There's a bios option to boot without keyboard, earlier mention as "halt on error" to disable and then you don't need a display also.
I didn't see anything of the like, and I've had other Linux distros running headless with no issues
 
Seems to be especially an issue if I have a VM configured to start on boot. I see the following in the log, could it have anything to do with proxmox dropping off the network?

Code:
Nov 13 21:45:23 vm kernel: tap100i0: entered promiscuous mode
Nov 13 21:45:23 vm kernel: vmbr0: port 2(fwpr100p0) entered blocking state
Nov 13 21:45:23 vm kernel: vmbr0: port 2(fwpr100p0) entered disabled state
Nov 13 21:45:23 vm kernel: fwpr100p0: entered allmulticast mode
Nov 13 21:45:23 vm kernel: fwpr100p0: entered promiscuous mode
Nov 13 21:45:23 vm kernel: vmbr0: port 2(fwpr100p0) entered blocking state
Nov 13 21:45:23 vm kernel: vmbr0: port 2(fwpr100p0) entered forwarding state
Nov 13 21:45:23 vm kernel: fwbr100i0: port 1(fwln100i0) entered blocking state
Nov 13 21:45:23 vm kernel: fwbr100i0: port 1(fwln100i0) entered disabled state
Nov 13 21:45:23 vm kernel: fwln100i0: entered allmulticast mode
Nov 13 21:45:23 vm kernel: fwln100i0: entered promiscuous mode
Nov 13 21:45:23 vm kernel: fwbr100i0: port 1(fwln100i0) entered blocking state
Nov 13 21:45:23 vm kernel: fwbr100i0: port 1(fwln100i0) entered forwarding state
Nov 13 21:45:23 vm kernel: fwbr100i0: port 2(tap100i0) entered blocking state
Nov 13 21:45:23 vm kernel: fwbr100i0: port 2(tap100i0) entered disabled state
Nov 13 21:45:23 vm kernel: tap100i0: entered allmulticast mode
Nov 13 21:45:23 vm kernel: fwbr100i0: port 2(tap100i0) entered blocking state
Nov 13 21:45:23 vm kernel: fwbr100i0: port 2(tap100i0) entered forwarding state

I currently do not have a display connected and did not start any VMs on boot or manually, and I am still able to access the web UI
 
So in fact you do have the static IP correctly setup within the Proxmox node. You should not need/have any reserved address within your router for this. Typically (as I said already above) you should have this address OUTSIDE the DHCP range of your router. Check for conflicts.

Seems to be especially an issue if I have a VM configured to start on boot
Do you have some type of passthrough going on within this VM?
 
Not an option, I do not have a display that I can dedicate to it. Other Linux OSs run just fine without a display or dummy adapter connected. If a hypervisor cannot run headless then something is wrong with it. I would venture to guess that MOST proxmox deployments are headless.

It's an Intel nuc 8thgen i3

Dude, you're trying to run on a NUC, not a dedicated 1U server with ILO/DRAC. NUCs are designed for desktop use.

You could probably do a run to your local pawn shop (or check FB marketplace, and other such-like websites for your local area) and grab a used monitor for ~$30. Otherwise, as others have mentioned you'll have to look up the motherboard manual and go through the BIOS settings to make sure it's not set to halt-on-error if it detects no kbd/monitor attached at boot. Lastly, your best/cheapest option is probably (as mentioned before) a dummy plug. Read the REFs below.

https://www.intel.com/content/www/us/en/support/articles/000091888/intel-nuc.html

https://www.reddit.com/r/homelab/comments/lfifca/possible_to_go_into_bios_when_trying_to_use_a_nuc/

https://www.reddit.com/r/intelnuc/comments/1cogza2/nuc8i58beh_as_headless_proxmox_host/
 
  • Like
Reactions: Johannes S
go through the BIOS settings to make sure it's not set to halt-on-error if it detects no kbd/monitor attached at boot
Don't think that is the problem since the OP writes:
My proxmox box boots up, if I'm quick I can log into the web UI, but then it will freeze and that's the end of it...
If it would BIOS halt-on-error, PVE would not boot up at all. See also above OP's post concerning VM at boot.
As I have indicated above, I suspect the problem is the NUC is loosing it's NW connection. Possibly this is caused by a reallocation of PCI devices within Proxmox that somehow is dependent on whether a device (monitor) is actually connected. Some power-saving feature? Some passthrough error? IDK.
 
Uups, I don't mean a cluster of rack servers, a single nuc is fine for a single pve at home usage but the old one has it's years coming to the end.
 
It doesn't make any sense to blame the hardware when proxmox is based on debian when normal debian and other distros run just fine on it, even headless. Must be something with the proxmox config

That being said, I'll probably just buy an HDMI emulator since it doesn't seem like we're moving towards a solution here. Hopefully there is a software solution though, as this is very strange and not something I've ever encountered with Linux before.
 
Last edited: