[WireGuard®](https://www.wireguard.com/) is an extremely simple yet fast and modern VPN that utilizes state-of-the-art cryptography. It aims to be faster, simpler, leaner, and more useful than IPsec, while avoiding the massive headache. It intends to be considerably more performant than OpenVPN. WireGuard is designed as a general purpose VPN for running on embedded interfaces and super computers alike, fit for many different circumstances. Initially released for the Linux kernel, it is now cross-platform (Windows, macOS, BSD, iOS, Android) and widely deployable. It is currently under heavy development, but already it might be regarded as the most secure, easiest to use, and simplest VPN solution in the industry.
Our images support multiple architectures such as `x86-64`, `arm64` and `armhf`. We utilise the docker manifest for multi-platform awareness. More information is available from docker [here](https://github.com/docker/distribution/blob/master/docs/spec/manifest-v2-2.md#manifest-list) and our announcement [here](https://blog.linuxserver.io/2019/02/21/the-lsio-pipeline-project/).
During container start, it will first check if the wireguard module is already installed and loaded. Kernels newer than 5.6 generally have the wireguard module built-in (along with some older custom kernels). However, the module may not be enabled. Make sure it is enabled prior to starting the container.
If the kernel is not built-in, or installed on host, the container will check if the kernel headers are present (in `/usr/src`) and if not, it will attempt to download the necessary kernel headers from the `ubuntu xenial/bionic`, `debian/raspbian buster` repos; then will attempt to compile and install the kernel module. If the kernel headers are not found in either `usr/src` or in the repos mentioned, container will sleep indefinitely as wireguard cannot be installed.
If you're on a debian/ubuntu based host with a custom or downstream distro provided kernel (ie. Pop!_OS), the container won't be able to install the kernel headers from the regular ubuntu and debian repos. In those cases, you can try installing the headers on the host via `sudo apt install linux-headers-$(uname -r)` (if distro version) and then add a volume mapping for `/usr/src:/usr/src`, or if custom built, map the location of the existing headers to allow the container to use host installed headers to build the kernel module (tested successful on Pop!_OS, ymmv).
With regards to arm32/64 devices, Raspberry Pi 2-4 running the [official ubuntu images](https://ubuntu.com/download/raspberry-pi) or Raspbian Buster are supported out of the box. For all other devices and OSes, you can try installing the kernel headers on the host, and mapping `/usr/src:/usr/src` and it may just work (no guarantees).
This can be run as a server or a client, based on the parameters used.
## Server Mode
If the environment variable `PEERS` is set to a number or a list of strings separated by comma, the container will run in server mode and the necessary server and peer/client confs will be generated. The peer/client config qr codes will be output in the docker log. They will also be saved in text and png format under `/config/peerX` in case `PEERS` is a variable and an integer or `/config/peer_X` in case a list of names was provided instead of an integer.
Variables `SERVERURL`, `SERVERPORT`, `INTERNAL_SUBNET` and `PEERDNS` are optional variables used for server mode. Any changes to these environment variables will trigger regeneration of server and peer confs. Peer/client confs will be recreated with existing private/public keys. Delete the peer folders for the keys to be recreated along with the confs.
To add more peers/clients later on, you increment the `PEERS` environment variable or add more elements to the list and recreate the container.
To display the QR codes of active peers again, you can use the following command and list the peer numbers as arguments: `docker exec -it wireguard /app/show-peer 1 4 5` or `docker exec -it wireguard /app/show-peer myPC myPhone myTablet` (Keep in mind that the QR codes are also stored as PNGs in the config folder).
The templates used for server and peer confs are saved under `/config/templates`. Advanced users can modify these templates and force conf generation by deleting `/config/wg0.conf` and restarting the container.
## Client Mode
Do not set the `PEERS` environment variable. Drop your client conf into the config folder as `/config/wg0.conf` and start the container.
If you get IPv6 related errors in the log and connection cannot be established, edit the `AllowedIPs` line in your peer/client wg0.conf to include only `0.0.0.0/0` and not `::/0`; and restart the container.
## Road warriors, roaming and returning home
If you plan to use Wireguard both remotely and locally, say on your mobile phone, you will need to consider routing. Most firewalls will not route ports forwarded on your WAN interface correctly to the LAN out of the box. This means that when you return home, even though you can see the Wireguard server, the return packets will probably get lost.
This is not a Wireguard specific issue and the two generally accepted solutions are NAT reflection (setting your edge router/firewall up in such a way as it translates internal packets correctly) or split horizon DNS (setting your internal DNS to return the private rather than public IP when connecting locally).
Both of these approaches have positives and negatives however their setup is out of scope for this document as everyone's network layout and equipment will be different.
## Maintaining local access to attached services
** Note: This is not a supported configuration by Linuxserver.io - use at your own risk.
When routing via Wireguard from another container using the `service` option in docker, you might lose access to the containers webUI locally. To avoid this, exclude the docker subnet from being routed via Wireguard by modifying your `wg0.conf` like so (modifying the subnets as you require):
```
[Interface]
PrivateKey = <privatekey>
Address = 9.8.7.6/32
DNS = 8.8.8.8
PostUp = DROUTE=$(ip route | grep default | awk '{print $3}'); HOMENET=192.168.0.0/16; HOMENET2=10.0.0.0/8; HOMENET3=172.16.0.0/12; ip route add $HOMENET3 via $DROUTE;ip route add $HOMENET2 via $DROUTE; ip route add $HOMENET via $DROUTE;iptables -I OUTPUT -d $HOMENET -j ACCEPT;iptables -A OUTPUT -d $HOMENET2 -j ACCEPT; iptables -A OUTPUT -d $HOMENET3 -j ACCEPT; iptables -A OUTPUT ! -o %i -m mark ! --mark $(wg show %i fwmark) -m addrtype ! --dst-type LOCAL -j REJECT
PreDown = HOMENET=192.168.0.0/16; HOMENET2=10.0.0.0/8; HOMENET3=172.16.0.0/12; ip route del $HOMENET3 via $DROUTE;ip route del $HOMENET2 via $DROUTE; ip route del $HOMENET via $DROUTE; iptables -D OUTPUT ! -o %i -m mark ! --mark $(wg show %i fwmark) -m addrtype ! --dst-type LOCAL -j REJECT; iptables -D OUTPUT -d $HOMENET -j ACCEPT; iptables -D OUTPUT -d $HOMENET2 -j ACCEPT; iptables -D OUTPUT -d $HOMENET3 -j ACCEPT
** Note: This is not a supported configuration by Linuxserver.io - use at your own risk.
Site-to-site VPN in server mode requires customizing the `AllowedIPs` statement for a specific peer in `wg0.conf`. Since `wg0.conf` is autogenerated when server vars are changed, it is not recommended to edit it manually.
In order to customize the `AllowedIPs` statement for a specific peer in `wg0.conf`, you can set an env var `SERVER_ALLOWEDIPS_PEER_<peer name or number>` to the additional subnets you'd like to add, comma separated and excluding the peer IP (ie. `"192.168.1.0/24,192.168.2.0/24"`). Replace `<peer name or number>` with either the name or number of a peer (whichever is used in the `PEERS` var).
For instance `SERVER_ALLOWEDIPS_PEER_laptop="192.168.1.0/24,192.168.2.0/24"` will result in the wg0.conf entry `AllowedIPs = 10.13.13.2,192.168.1.0/24,192.168.2.0/24` for the peer named `laptop`.
Keep in mind that this var will only be considered when the confs are regenerated. Adding this var for an existing peer won't force a regeneration. You can delete wg0.conf and restart the container to force regeneration if necessary.
Don't forget to set the necessary POSTUP and POSTDOWN rules in your client's peer conf for lan access.
Docker images are configured using parameters passed at runtime (such as those above). These parameters are separated by a colon and indicate `<external>:<internal>` respectively. For example, `-p 8080:80` would expose port `80` from inside the container to be accessible from the host's IP on port `8080` outside the container.
| `SERVERURL=wireguard.domain.com` | External IP or domain name for docker host. Used in server mode. If set to `auto`, the container will try to determine and set the external IP automatically |
| `SERVERPORT=51820` | External port for docker host. Used in server mode. |
| `PEERS=1` | Number of peers to create confs for. Required for server mode. Can also be a list of names: `myPC,myPhone,myTablet` (alphanumeric only) |
| `PEERDNS=auto` | DNS server set in peer/client configs (can be set as `8.8.8.8`). Used in server mode. Defaults to `auto`, which uses wireguard docker host's DNS via included CoreDNS forward. |
| `INTERNAL_SUBNET=10.13.13.0` | Internal subnet for the wireguard and server and peers (only change if it clashes). Used in server mode. |
| `ALLOWEDIPS=0.0.0.0/0` | The IPs/Ranges that the peers will be able to reach using the VPN connection. If not specified the default value is: '0.0.0.0/0, ::0/0' This will cause ALL traffic to route through the VPN, if you want split tunneling, set this to only the IPs you would like to use the tunnel AND the ip of the server's WG ip, such as 10.13.13.1. |
For all of our images we provide the ability to override the default umask settings for services started within the containers using the optional `-e UMASK=022` setting.
Keep in mind umask is not chmod it subtracts from permissions based on it's value it does not add. Please read up [here](https://en.wikipedia.org/wiki/Umask) before asking for support.
When using volumes (`-v` flags), permissions issues can arise between the host OS and the container, we avoid this issue by allowing you to specify the user `PUID` and group `PGID`.
[](https://mods.linuxserver.io/?mod=wireguard "view available mods for this container.") [](https://mods.linuxserver.io/?mod=universal "view available universal mods.")
We publish various [Docker Mods](https://github.com/linuxserver/docker-mods) to enable additional functionality within the containers. The list of Mods available for this image (if any) as well as universal mods that can be applied to any one of our images can be accessed via the dynamic badges above.
* **23.04.22:** - Add pre-shared key support. Automatically added to all new peer confs generated, existing ones are left without to ensure no breaking changes.
* **04.10.20:** - Allow to specify a list of names as PEERS and add ALLOWEDIPS environment variable. Also, add peer name/id to each one of the peer sections in wg0.conf. Important: Existing users need to delete `/config/templates/peer.conf` and restart
* **06.08.20:** - Replace resolvconf with openresolv due to dns issues when a client based on this image is connected to a server also based on this image. Add IPv6 info to readme. Display kernel version in logs.
* **29.07.20:** - Update Coredns config to detect dns loops (existing users need to delete `/config/coredns/Corefile` and restart).
* **27.07.20:** - Update Coredns config to prevent issues with non-user-defined bridge networks (existing users need to delete `/config/coredns/Corefile` and restart).
* **19.06.20:** - Add support for Ubuntu Focal (20.04) kernels. Compile wireguard tools and kernel module instead of using the ubuntu packages. Make module install optional. Improve verbosity in logs.
* **08.04.20:** - Add arm32/64 builds and enable multi-arch (rpi4 with ubuntu and raspbian buster tested). Add CoreDNS for `PEERDNS=auto` setting. Update the `add-peer`/`show-peer` scripts to utilize the templates and the `INTERNAL_SUBNET` var (previously missed, oops).
