Unable to connect to web interface or SSH, but ping works

jrmoulckers

New Member
Nov 7, 2025
1
0
1
TL;DR
I have been struggling with my fresh Proxmox installation on a dedicated server from my Windows workstation when trying to connect via SSH or Web GUI.

I have tried all of the checks and solutions I have seen so far on other guides here, and tried to get Copilot/ChatGPT to help me (just gave generic solutions).

Quite new to PVE/Linux world, but I have some exposure thru tech/homelabbing.

Bottom line:
  • I am only able to access PVE physically through its host console
  • I can ping PVE at its 192.168.4.111 from my workstation
  • I can ping my workstation at its 192.168.4.222 from PVE
  • Both commands return HTML executing on PVE host:
    • curl -k https://192.168.4.111:8006/
    • curl -k https://localhost:8006/
  • I cannot access the web interface at https://192.168.4.111:8006/ from my workstation
  • SSH hangs from my workstation: ssh root@192.168.4.111
Troubleshooting (Server):
  • ip a output:1762558164155.jpeg
  • ifdown vmbr0 and ifup vmbr0 does not change the interface's IP
  • /etc/network/interfaces contents:1762558375657.jpeg
  • /etc/hosts contents:1762559435941.jpeg
  • I have tried changing the network mask to /24, despite my Workstation displaying /22, but this did not affect anything after restarting networking and reboot
  • (As above) PVE receives HTML when running curl at either IP or localhost
  • I've tried stopping and disabling the firewall: pve-firewall stop systemctl disable pve-firewall
  • I've tried adding the file /etc/default/pveproxywith the following contents:
    • LISTEN 0.0.0.0
  • Every file change I have restarted the respective service and/or the host
  • PVE is connected to internet, and can ping 8.8.8.8 and curl google.com, etc.
  • nmap -p 8006 192.168.4.111 output:1762559454518.jpeg
  • I have tried restarting both pveproxy, pve-cluster and pvedaemon services
  • journalctl -f shows no errors or warnings, and shows proper service startup
  • I have also tried some commands like ss, with grep on port 8006, and I can see pings and acks during ssh and webpage requests, so I feel like the ack is not making it back to my workstation??
Troubleshooting (Workstation):
  • (As above) Ping works, SSH and Web GUI do not
  • curlhangs and fails
  • CLI connection results:
Test-NetConnection -ComputerName 192.168.4.111 -Port 8006
ComputerName : 192.168.4.111
RemoteAddress : 192.168.4.111
RemotePort : 8006
NameResolutionResults : 192.168.4.111
MatchingIPsecRules :
NetworkIsolationContext : Private Network
IsAdmin : False
InterfaceAlias : Ethernet 2
SourceAddress : 192.168.4.222
NetRoute (NextHop) : 192.168.4.1
TcpTestSucceeded : True
  • There were some issues with being unable to ping the server before, but I de-prioritized Tailscale's interface metric (see Network Details below), which resolved the conflicting interface
  • As far as I can tell, there are no outbound firewall rules blocking anything from Windows
Troubleshooting (Router):
  • I have restarted my network multiple times
  • I have tried unassigning IP addresses, removing reservations, clearing them by resetting, etc.

Here is my hardware setup:
  • Server: Proxmox 9.0.3 on dedicated Dell Precision Tower
  • Workstation: Windows 11 on another rig
  • Network: Eero 6 (EeroOS 7.x.x)
Network details:
I have both my server and workstation wired into one of the Eero units, which is connected wirelessly as an extender of the main Eero router (if this matters)
I have Eero reserving IP addresses for each device:
Server: 192.168.4.111
Workstation: 192.168.4.222
I have Tailscale installed on my workstation, but I have an InterfaceMetric defined to de-prioritize it under my active Ethernet interface:
Code:
InterfaceAlias    InterfaceMetric
Ethernet 2         5
...
Tailscale 50

My Ideas:
Eero may be treating my server as a public route???
Eero could be blocking traffic to a suspicious port (8006) without SSL???

I have been able to access the GUI before, but as soon as I started reserving IP addresses on the physical hosts - I had already reserved successfully for VMs and LXCs - and made a fresh installation, this issue came up.

Thanks in advance! Tried to be as thorough as possible to cover all bases
 
Last edited: