Virtual Box与亚马逊实例通信报错：Python socket.error: Connection refused

Let’s work through the common issues that could be causing this connection failure—since you’ve already covered the basics (port opening, DMZ setup, bridge mode, firewall disabled), we’ll dig into the less obvious checks:

• Confirm your VM’s listener is bound to the correct address
  It’s incredibly common for socket programs to default to listening only on 127.0.0.1 (localhost), which means they’ll only accept connections from inside the VM, not external sources like your AWS instance.
  Run this command on your Ubuntu VM to verify:

  ss -tulpn | grep 2019
  

  Look for a line with LISTEN in it. The address column should show 0.0.0.0:2019 (accepts connections from any IP) or your VM’s LAN IP (e.g., 192.168.1.100:2019). If it shows 127.0.0.1:2019, modify your Python code to bind to 0.0.0.0 instead of localhost.

• Double-check your VM’s LAN IP matches your DMZ configuration
  Bridge mode can assign a new IP if your router’s DHCP lease expires. Confirm your VM’s current IP with:

  ip a
  

  Compare this to the IP you set in your router’s DMZ settings. If they don’t match, update the DMZ to use the correct active IP.

• Test connectivity from your AWS instance to your public IP
  First, get your home network’s public IP (you can find this in your router’s admin panel or by running a public IP lookup from a device on your LAN). Then from your AWS DL AMI instance, run:

  telnet <your-public-ip> 2019
  

  If this fails, either your ISP is blocking incoming traffic on port 2019, or your DMZ/port forwarding isn’t properly configured. If telnet works but your Python code doesn’t, you likely have a protocol mismatch (TCP vs UDP) in your code.

• Verify VirtualBox’s bridge mode uses the right physical adapter
  If your host machine has multiple adapters (e.g., Wi-Fi and Ethernet), VirtualBox might be bridging to the wrong one. Open VirtualBox, go to your VM’s Settings > Network > Adapter 1, ensure "Bridge Adapter" is selected, and the dropdown matches the adapter connected to your LAN (the one your router uses for wired/wireless traffic).

• Check for residual iptables rules (even with ufw disabled)
  Ubuntu can sometimes leave leftover iptables rules after turning off ufw. List all rules with:

  sudo iptables -L -n
  

  Look for any REJECT or DROP rules targeting port 2019. If you find them, flush the rules with:

  sudo iptables -F
  

• Rule out protocol mismatch (TCP vs UDP)
  You mentioned opening both protocols on port 2019, but make sure your AWS client uses the same protocol as your VM’s server:

  • If your VM runs a TCP listener, your AWS client must use socket.SOCK_STREAM, not SOCK_DGRAM.
  • UDP doesn’t use a "connection" step, so a "connection refused" error here almost always means you’re mixing TCP and UDP in your client/server code.

内容的提问来源于stack exchange，提问作者Alberto Fernández