I'm encountering an issue while trying to enable the "Host access to custom networks" feature in Docker on my Unraid system. Despite configuring it according to the documentation, the setting fails to work as expected. I've captured some key logs that might shed light on the problem and I'll also include the relevant network configuration.

 

System Information

OS: Unraid 6.12.9

 

Key Logs

Nov 14 10:37:07 AIO root: Error response from daemon: Pool overlaps with other one on this address space
Nov 14 10:37:07 AIO root: RTNETLINK answers: Device or resource busy
Nov 14 10:37:07 AIO kernel: br0: Device is already in use.
Nov 14 10:37:07 AIO root: Device "shim-br0" does not exist.
Nov 14 10:37:07 AIO root: Cannot find device "shim-br0"
Nov 14 10:37:07 AIO root: /etc/rc.d/rc.docker: line 417: /proc/sys/net/ipv6/conf/shim-br0/disable_ipv6: No such file or directory
Nov 14 10:37:07 AIO root: Cannot find device "shim-br0"
Nov 14 10:37:07 AIO root: Cannot find device "shim-br0"
Nov 14 10:37:07 AIO rc.docker: created network shim-br0 for host access

 

Docker NetWorks

NETWORK ID     NAME                           DRIVER    SCOPE
.......
49896d032498   docker_vlan                    macvlan   local
c4d14cdf4426   host                           host      local
9b4eb889b406   none                           null      local

version: '3.3'
services:
  .........
  qbittorrent:
    image: 'linuxserver/qbittorrent:4.6.3'
    container_name: qbittorrent
    restart: always
    hostname: qbittorrent
    blkio_config:
       weight: 100
    volumes:
        - '/mnt/user/appdata/qb/config:/config'
        - '/mnt/user/data/download:/download'
    tmpfs:
        - '/tmp'
    networks:
        docker_vlan:
          ipv4_address: 192.168.31.4

networks:
  docker_vlan: # 和宿主机同一个网络的情况下，只能创建这一个了
    name: docker_vlan
    driver: macvlan
    driver_opts:
      parent: br0  # 使用的主机网络接口
    ipam:
      config:
        - subnet: 192.168.31.0/24
          gateway: 192.168.31.1

part of docker compose

Host Network

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
       valid_lft forever preferred_lft forever
2: tunl0@NONE: <NOARP> mtu 1480 qdisc noop state DOWN group default qlen 1000
    link/ipip 0.0.0.0 brd 0.0.0.0
6: eth0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc fq master bond0 state UP group default qlen 1000
    link/ether 84:47:09:00:06:b4 brd ff:ff:ff:ff:ff:ff
7: eth1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc fq master bond0 state UP group default qlen 1000
    link/ether 84:47:09:00:06:b4 brd ff:ff:ff:ff:ff:ff permaddr 84:47:09:00:06:b5
8: eth2: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc fq master bond0 state UP group default qlen 1000
    link/ether 84:47:09:00:06:b4 brd ff:ff:ff:ff:ff:ff permaddr 84:47:09:00:06:b6
9: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue master br0 state UP group default qlen 1000
    link/ether 84:47:09:00:06:b4 brd ff:ff:ff:ff:ff:ff
10: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
    link/ether 84:47:09:00:06:b4 brd ff:ff:ff:ff:ff:ff
    inet 192.168.31.170/24 brd 192.168.31.255 scope global dynamic noprefixroute br0
       valid_lft 46690sec preferred_lft 35890sec
 ..........

 

update to lattestest stable.

since macvlan. turn off bridging to test...
I'm not aware of Unraid 6.12.9 having the macvlan patch. I though that was 6.12.12 / 6.12.13
*Need to find the release note on it again...

Please post diag file

I've made some progress since my last post and wanted to share what I found, but I'm still facing some challenges and could really use some help.

What I've Discovered

After investigating the issue further, I realized that the "Host Access to Custom Networks" feature essentially creates a separate macvlan interface to facilitate communication between the host and Docker's macvlan.

To work around my original issue, I manually created a vlan_bridge0 interface, and this allowed my Unraid host to communicate with the Docker macvlan network. As a result, I can now see the container-assigned IP addresses on my router, and they are accessible as expected.

New Problem with WireGuard

However, I'm encountering a new issue related to WireGuard.

Here’s what I’ve done:

Disabled NAT in WireGuard as per the guide.

Added a static route on my router for the WireGuard subnet to route through the Unraid server.

I can successfully ping devices through WireGuard, including those on the macvlan network.

The problem: When I try to access web pages (HTTP) on the macvlan network through WireGuard, the connection is extremely slow—almost unusable. In contrast:

Accessing other LAN IPs through WireGuard works perfectly and at normal speeds.

Accessing the macvlan network from other LAN devices (not using WireGuard) is also fast and responsive.

I'm unsure what could be causing this bottleneck when accessing the macvlan network through WireGuard, especially since ping works fine and other traffic is unaffected. Has anyone else experienced something similar or can offer advice on how to troubleshoot this further?

Thanks again for your help!

 

ip link add vlan_bridge0 link br0 type macvlan mode bridge
ip addr add 192.168.31.2 dev vlan_bridge0
ip link set vlan_bridge0 up
ip route add 192.168.31.4 dev vlan_bridge0

1 minute ago, bmartino1 said:

update to lattestest stable.

since macvlan. turn off bridging to test...
I'm not aware of Unraid 6.12.9 having the macvlan patch. I though that was 6.12.12 / 6.12.13
*Need to find the release note on it again...

Please post diag file

thanks reply, i will update os in a while

The Realtek RTL8125 network card is used in conjunction with the network card driver, and there may be some issues with the default driver of unraid. As long as the 8125 driver is installed, this problem will be solved