Alright, let's tackle this problem head-on. The core issue here is that your organization's internal network is using Docker's default 172.17.0.1 gateway, so you need to reconfigure the docker0 bridge to use a custom subnet—but the default-gateway setting is causing Docker to fail when set to 192.168.220.1. Here's how to resolve this:

• Diagnose the Current Configuration

  • First, check your existing Docker daemon config and network setup:
    • View your current daemon.json:

      cat /etc/docker/daemon.json
      

    • Check if 192.168.220.1 is already in use by another network interface on your host:

      ip addr show
      

  • Common failure causes include: the gateway IP is already occupied by another NIC (like eth0 or wlan0), or your subnet definition doesn't include the gateway IP.

• Correctly Configure the Custom Subnet

  • The bip (bridge IP) parameter in daemon.json defines both the docker0 subnet and its gateway in one line, avoiding conflicts with a separate default-gateway setting. Use this valid config:

    {
      "bip": "192.168.220.1/24"
    }
    

  • This sets docker0 to use the 192.168.220.0/24 subnet with 192.168.220.1 as the gateway. If that IP is taken, pick another valid one in the subnet (e.g., 192.168.220.254) and adjust the bip:

    {
      "bip": "192.168.220.254/24"
    }
    

• Clean Up and Restart Docker

  • Before restarting, remove the old docker0 bridge to avoid conflicts:
    1. Stop the Docker service:

       systemctl stop docker
       

    2. Install bridge utilities if missing:

       apt update && apt install bridge-utils -y
       

    3. Delete the existing docker0 bridge:

       ip link set dev docker0 down && brctl delbr docker0
       

    4. Start Docker back up:

       systemctl start docker
       

• Verify the Fix

  • Check if docker0 is using your new subnet:

    ip addr show docker0
    

  • Test internet access from a container:

    docker run --rm alpine ping -c 3 google.com
    

  • If the ping succeeds, your containers now have internet access.

• Troubleshoot Persistent Startup Failures

  • If Docker still won't start, check service logs for specific errors:

    journalctl -u docker.service -f
    

  • Logs will reveal issues like IP conflicts, invalid subnet ranges, or other configuration problems to adjust.

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