hotio/nzbhydra2

Starting the container¶

clicompose

docker run --rm \
    --name nzbhydra2 \
    -p 5076:5076 \
    -e PUID=1000 \
    -e PGID=1000 \
    -e UMASK=002 \
    -e TZ="Etc/UTC" \
    -v /<host_folder_config>:/config \
    ghcr.io/hotio/nzbhydra2

services:
  nzbhydra2:
    container_name: nzbhydra2
    image: ghcr.io/hotio/nzbhydra2
    ports:
      - "5076:5076"
    environment:
      - PUID=1000
      - PGID=1000
      - UMASK=002
      - TZ=Etc/UTC
    volumes:
      - /<host_folder_config>:/config

Tags¶

Tags	Description	Last Updated	Age

WireGuard¶

Info

This image includes VPN support. The cli/compose examples below are environment variables and settings complementary to the app image examples, this means you'll have to add/merge the stuff below with the stuff above. In case you are still in need of a VPN, consider using my affiliate link for Proton or Private Internet Access.

genericprotonpia

clicompose

docker run --rm \
    --hostname container-name.internal \ #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    -e VPN_ENABLED="true" \ #
There needs to be a file wg0.conf located in /config/wireguard and you need to set the variable VPN_ENABLED to true for the VPN to start. If you'd like to execute some of your own bash commands you can place two scripts alongside your wg0.conf, called wg0-pre.sh and wg0-post.sh. These will execute right before the check for the existence of a wg0.conf file and almost at the end, right before the internet connectivity test.

    -e VPN_CONF="wg0" \ # READ THIS
With VPN_CONF you can set the name used for your WireGuard config. This is an example of how your wg0.conf file should look like. If there's a lot of extra stuff, remove it unless you know what it's there for. The WireGuard config is automatically modified to use AllowedIPs = 0.0.0.0/1,128.0.0.0/1 for compatibility with Synology/QNAP/Asustor/WSL2 systems if you append -fix, so VPN_CONF=wg0-fix. WSL2 users can also compile their own kernel if they don't wanna use this workaround.
[Interface]
PrivateKey = supersecretprivatekey
Address = xx.xx.xxx.xxx/32 # Yes, /32
DNS = x.x.x.x

[Peer]
PublicKey = publickey
AllowedIPs = 0.0.0.0/0
Endpoint = xxx.x.xxx.x:51820


    -e VPN_PROVIDER="generic" \ #
Possible values are generic, proton and pia.

    -e VPN_LAN_NETWORK="192.168.1.0/24" \ #
The environment variable VPN_LAN_NETWORK can be set to for example 192.168.1.0/24, 192.168.1.0/24,192.168.44.0/24 or 192.168.1.33, so you can get access to the webui or other additional ports (see below). If for example you were to pick 192.168.0.0/24, every device with an ip in the range 192.168.0.0 - 192.168.0.255 on your LAN is allowed access to the webui.

    -e VPN_LAN_LEAK_ENABLED="false" \
    -e VPN_EXPOSE_PORTS_ON_LAN="" \ #
If you need to expose ports on your LAN you can use VPN_EXPOSE_PORTS_ON_LAN. For example VPN_EXPOSE_PORTS_ON_LAN=7878/tcp,9117/tcp, will block those ports on the vpn interface, so that there's no risk that they might be exposed to the world and allow access to them from your LAN. Some images also have a WEBUI_PORTS environment variable that does basically the same for the vpn part. For those apps that support it, it'll also change the port on which the app runs.

    -e VPN_AUTO_PORT_FORWARD="false" \ #
Auto retrieve a forwarded port and configure the supported app if set to true or if you can manually request/set a forwarded port in the VPN provider's web interface, fill in the port number (just the number). Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

    -e VPN_AUTO_PORT_FORWARD_TO_PORTS="" \ #
Adds a redirect for the forwarded port from your vpn provider to the internal port on which the app runs, ports in this list are also not blocked on the wireguard interface, so this var is also useful if you want to expose a port on both your LAN and VPN. Values like 32400/tcp will use the port from VPN_AUTO_PORT_FORWARD to create the redirect or if set to true the forwarded port from pia/proton. Use 3000@3001/tcp,3002@3003/tcp syntax for extra static redirects. The only known usecase as of right now is Plex and exposing it on the VPN with a non configurable forwarded port, because it's not possible to run Plex on anything else but 32400. Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

    -e VPN_KEEP_LOCAL_DNS="false" \ #
Setting this to true will re-add the default nameserver if it's been overwritten by the DNS = ... entry in wgO.conf. This should keep container name resolution working. If container name resolution still doesn't work, add --dns 1.1.1.1, somehow this can fix it (confirmed on MacOS).

    -e VPN_FIREWALL_TYPE="auto" \ #
Possible values are auto, legacy or nftables. The default is auto, this will try to use the most modern method available. If this doesn't work, you can try forcing it to legacy or nftables.

    -e VPN_HEALTHCHECK_ENABLED="false" \
    -e PRIVOXY_ENABLED="false" \
    -e UNBOUND_ENABLED="false" \ #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    --cap-add=NET_ADMIN \
    --sysctl="net.ipv4.conf.all.src_valid_mark=1" \ #
Required in most cases, on some systems that don't have rp_filter set to strict, it's optional.

    --sysctl="net.ipv6.conf.all.disable_ipv6=1" \ #
With net.ipv6.conf.all.disable_ipv6=1 all ipv6 support is disabled. Leave ipv6 disabled and remove all references to it in your wg0.conf file to keep things simple. If you need any sort of ipv6 support, enable it with net.ipv6.conf.all.disable_ipv6=0. A WireGuard ipv6 endpoint is currently not supported.

    --device /dev/net/tun:/dev/net/tun \ # OPTIONAL, READ THIS
If the WireGuard kernel module is missing (most likely on Synology/QNAP/Asustor), you can run WireGuard in userspace thanks to wireguard-go. For that you'll need to add the device /dev/net/tun. It's most likely that the device /dev/net/tun does not exist however, have a read here for instructions on checking and adding the device.

    ...

services:
  app:
    hostname: container-name.internal #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    environment:
      - VPN_ENABLED=true #
There needs to be a file wg0.conf located in /config/wireguard and you need to set the variable VPN_ENABLED to true for the VPN to start. If you'd like to execute some of your own bash commands you can place two scripts alongside your wg0.conf, called wg0-pre.sh and wg0-post.sh. These will execute right before the check for the existence of a wg0.conf file and almost at the end, right before the internet connectivity test.

      - VPN_CONF=wg0 # READ THIS
With VPN_CONF you can set the name used for your WireGuard config. This is an example of how your wg0.conf file should look like. If there's a lot of extra stuff, remove it unless you know what it's there for. The WireGuard config is automatically modified to use AllowedIPs = 0.0.0.0/1,128.0.0.0/1 for compatibility with Synology/QNAP/Asustor/WSL2 systems if you append -fix, so VPN_CONF=wg0-fix. WSL2 users can also compile their own kernel if they don't wanna use this workaround.
[Interface]
PrivateKey = supersecretprivatekey
Address = xx.xx.xxx.xxx/32 # Yes, /32
DNS = x.x.x.x

[Peer]
PublicKey = publickey
AllowedIPs = 0.0.0.0/0
Endpoint = xxx.x.xxx.x:51820


      - VPN_PROVIDER=generic #
Possible values are generic, proton and pia.

      - VPN_LAN_NETWORK=192.168.1.0/24 #
The environment variable VPN_LAN_NETWORK can be set to for example 192.168.1.0/24, 192.168.1.0/24,192.168.44.0/24 or 192.168.1.33, so you can get access to the webui or other additional ports (see below). If for example you were to pick 192.168.0.0/24, every device with an ip in the range 192.168.0.0 - 192.168.0.255 on your LAN is allowed access to the webui.

      - VPN_LAN_LEAK_ENABLED=false
      - VPN_EXPOSE_PORTS_ON_LAN #
If you need to expose ports on your LAN you can use VPN_EXPOSE_PORTS_ON_LAN. For example VPN_EXPOSE_PORTS_ON_LAN=7878/tcp,9117/tcp, will block those ports on the vpn interface, so that there's no risk that they might be exposed to the world and allow access to them from your LAN. Some images also have a WEBUI_PORTS environment variable that does basically the same for the vpn part. For those apps that support it, it'll also change the port on which the app runs.

      - VPN_AUTO_PORT_FORWARD=false #
Auto retrieve a forwarded port and configure the supported app if set to true or if you can manually request/set a forwarded port in the VPN provider's web interface, fill in the port number (just the number). Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

      - VPN_AUTO_PORT_FORWARD_TO_PORTS= #
Adds a redirect for the forwarded port from your vpn provider to the internal port on which the app runs, ports in this list are also not blocked on the wireguard interface, so this var is also useful if you want to expose a port on both your LAN and VPN. Values like 32400/tcp will use the port from VPN_AUTO_PORT_FORWARD to create the redirect or if set to true the forwarded port from pia/proton. Use 3000@3001/tcp,3002@3003/tcp syntax for extra static redirects. The only known usecase as of right now is Plex and exposing it on the VPN with a non configurable forwarded port, because it's not possible to run Plex on anything else but 32400. Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

      - VPN_KEEP_LOCAL_DNS=false #
Setting this to true will re-add the default nameserver if it's been overwritten by the DNS = ... entry in wgO.conf. This should keep container name resolution working. If container name resolution still doesn't work, add --dns 1.1.1.1, somehow this can fix it (confirmed on MacOS).

      - VPN_FIREWALL_TYPE=auto #
Possible values are auto, legacy or nftables. The default is auto, this will try to use the most modern method available. If this doesn't work, you can try forcing it to legacy or nftables.

      - VPN_HEALTHCHECK_ENABLED=false
      - PRIVOXY_ENABLED=false
      - UNBOUND_ENABLED=false #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    cap_add:
      - NET_ADMIN
    sysctls:
      - net.ipv4.conf.all.src_valid_mark=1 #
Required in most cases, on some systems that don't have rp_filter set to strict, it's optional.

      - net.ipv6.conf.all.disable_ipv6=1 #
With net.ipv6.conf.all.disable_ipv6=1 all ipv6 support is disabled. Leave ipv6 disabled and remove all references to it in your wg0.conf file to keep things simple. If you need any sort of ipv6 support, enable it with net.ipv6.conf.all.disable_ipv6=0. A WireGuard ipv6 endpoint is currently not supported.

    devices:
      - /dev/net/tun:/dev/net/tun # OPTIONAL, READ THIS
If the WireGuard kernel module is missing (most likely on Synology/QNAP/Asustor), you can run WireGuard in userspace thanks to wireguard-go. For that you'll need to add the device /dev/net/tun. It's most likely that the device /dev/net/tun does not exist however, have a read here for instructions on checking and adding the device.

    ...

clicompose

docker run --rm \
    --hostname container-name.internal \ #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    -e VPN_ENABLED="true" \ #
There needs to be a file wg0.conf located in /config/wireguard and you need to set the variable VPN_ENABLED to true for the VPN to start. If you'd like to execute some of your own bash commands you can place two scripts alongside your wg0.conf, called wg0-pre.sh and wg0-post.sh. These will execute right before the check for the existence of a wg0.conf file and almost at the end, right before the internet connectivity test.

    -e VPN_CONF="wg0" \ # READ THIS
With VPN_CONF you can set the name used for your WireGuard config. This is an example of how your wg0.conf file should look like. If there's a lot of extra stuff, remove it unless you know what it's there for. The WireGuard config is automatically modified to use AllowedIPs = 0.0.0.0/1,128.0.0.0/1 for compatibility with Synology/QNAP/Asustor/WSL2 systems if you append -fix, so VPN_CONF=wg0-fix. WSL2 users can also compile their own kernel if they don't wanna use this workaround.
[Interface]
PrivateKey = supersecretprivatekey
Address = xx.xx.xxx.xxx/32 # Yes, /32
DNS = x.x.x.x

[Peer]
PublicKey = publickey
AllowedIPs = 0.0.0.0/0
Endpoint = xxx.x.xxx.x:51820


    -e VPN_PROVIDER="proton" \ #
Possible values are generic, proton and pia.

    -e VPN_LAN_NETWORK="192.168.1.0/24" \ #
The environment variable VPN_LAN_NETWORK can be set to for example 192.168.1.0/24, 192.168.1.0/24,192.168.44.0/24 or 192.168.1.33, so you can get access to the webui or other additional ports (see below). If for example you were to pick 192.168.0.0/24, every device with an ip in the range 192.168.0.0 - 192.168.0.255 on your LAN is allowed access to the webui.

    -e VPN_LAN_LEAK_ENABLED="false" \
    -e VPN_EXPOSE_PORTS_ON_LAN="" \ #
If you need to expose ports on your LAN you can use VPN_EXPOSE_PORTS_ON_LAN. For example VPN_EXPOSE_PORTS_ON_LAN=7878/tcp,9117/tcp, will block those ports on the vpn interface, so that there's no risk that they might be exposed to the world and allow access to them from your LAN. Some images also have a WEBUI_PORTS environment variable that does basically the same for the vpn part. For those apps that support it, it'll also change the port on which the app runs.

    -e VPN_AUTO_PORT_FORWARD="true" \ #
Auto retrieve a forwarded port and configure the supported app if set to true or if you can manually request/set a forwarded port in the VPN provider's web interface, fill in the port number (just the number). Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

    -e VPN_AUTO_PORT_FORWARD_TO_PORTS="" \ #
Adds a redirect for the forwarded port from your vpn provider to the internal port on which the app runs, ports in this list are also not blocked on the wireguard interface, so this var is also useful if you want to expose a port on both your LAN and VPN. Values like 32400/tcp will use the port from VPN_AUTO_PORT_FORWARD to create the redirect or if set to true the forwarded port from pia/proton. Use 3000@3001/tcp,3002@3003/tcp syntax for extra static redirects. The only known usecase as of right now is Plex and exposing it on the VPN with a non configurable forwarded port, because it's not possible to run Plex on anything else but 32400. Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

    -e VPN_KEEP_LOCAL_DNS="false" \ #
Setting this to true will re-add the default nameserver if it's been overwritten by the DNS = ... entry in wgO.conf. This should keep container name resolution working. If container name resolution still doesn't work, add --dns 1.1.1.1, somehow this can fix it (confirmed on MacOS).

    -e VPN_FIREWALL_TYPE="auto" \ #
Possible values are auto, legacy or nftables. The default is auto, this will try to use the most modern method available. If this doesn't work, you can try forcing it to legacy or nftables.

    -e VPN_HEALTHCHECK_ENABLED="false" \
    -e PRIVOXY_ENABLED="false" \
    -e UNBOUND_ENABLED="false" \ #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    --cap-add=NET_ADMIN \
    --sysctl="net.ipv4.conf.all.src_valid_mark=1" \ #
Required in most cases, on some systems that don't have rp_filter set to strict, it's optional.

    --sysctl="net.ipv6.conf.all.disable_ipv6=1" \ #
With net.ipv6.conf.all.disable_ipv6=1 all ipv6 support is disabled. Leave ipv6 disabled and remove all references to it in your wg0.conf file to keep things simple. If you need any sort of ipv6 support, enable it with net.ipv6.conf.all.disable_ipv6=0. A WireGuard ipv6 endpoint is currently not supported.

    --device /dev/net/tun:/dev/net/tun \ # OPTIONAL, READ THIS
If the WireGuard kernel module is missing (most likely on Synology/QNAP/Asustor), you can run WireGuard in userspace thanks to wireguard-go. For that you'll need to add the device /dev/net/tun. It's most likely that the device /dev/net/tun does not exist however, have a read here for instructions on checking and adding the device.

    ...

services:
  app:
    hostname: container-name.internal #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    environment:
      - VPN_ENABLED=true #
There needs to be a file wg0.conf located in /config/wireguard and you need to set the variable VPN_ENABLED to true for the VPN to start. If you'd like to execute some of your own bash commands you can place two scripts alongside your wg0.conf, called wg0-pre.sh and wg0-post.sh. These will execute right before the check for the existence of a wg0.conf file and almost at the end, right before the internet connectivity test.

      - VPN_CONF=wg0 # READ THIS
With VPN_CONF you can set the name used for your WireGuard config. This is an example of how your wg0.conf file should look like. If there's a lot of extra stuff, remove it unless you know what it's there for. The WireGuard config is automatically modified to use AllowedIPs = 0.0.0.0/1,128.0.0.0/1 for compatibility with Synology/QNAP/Asustor/WSL2 systems if you append -fix, so VPN_CONF=wg0-fix. WSL2 users can also compile their own kernel if they don't wanna use this workaround.
[Interface]
PrivateKey = supersecretprivatekey
Address = xx.xx.xxx.xxx/32 # Yes, /32
DNS = x.x.x.x

[Peer]
PublicKey = publickey
AllowedIPs = 0.0.0.0/0
Endpoint = xxx.x.xxx.x:51820


      - VPN_PROVIDER=proton #
Possible values are generic, proton and pia.

      - VPN_LAN_NETWORK=192.168.1.0/24 #
The environment variable VPN_LAN_NETWORK can be set to for example 192.168.1.0/24, 192.168.1.0/24,192.168.44.0/24 or 192.168.1.33, so you can get access to the webui or other additional ports (see below). If for example you were to pick 192.168.0.0/24, every device with an ip in the range 192.168.0.0 - 192.168.0.255 on your LAN is allowed access to the webui.

      - VPN_LAN_LEAK_ENABLED=false
      - VPN_EXPOSE_PORTS_ON_LAN #
If you need to expose ports on your LAN you can use VPN_EXPOSE_PORTS_ON_LAN. For example VPN_EXPOSE_PORTS_ON_LAN=7878/tcp,9117/tcp, will block those ports on the vpn interface, so that there's no risk that they might be exposed to the world and allow access to them from your LAN. Some images also have a WEBUI_PORTS environment variable that does basically the same for the vpn part. For those apps that support it, it'll also change the port on which the app runs.

      - VPN_AUTO_PORT_FORWARD=true #
Auto retrieve a forwarded port and configure the supported app if set to true or if you can manually request/set a forwarded port in the VPN provider's web interface, fill in the port number (just the number). Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

      - VPN_AUTO_PORT_FORWARD_TO_PORTS= #
Adds a redirect for the forwarded port from your vpn provider to the internal port on which the app runs, ports in this list are also not blocked on the wireguard interface, so this var is also useful if you want to expose a port on both your LAN and VPN. Values like 32400/tcp will use the port from VPN_AUTO_PORT_FORWARD to create the redirect or if set to true the forwarded port from pia/proton. Use 3000@3001/tcp,3002@3003/tcp syntax for extra static redirects. The only known usecase as of right now is Plex and exposing it on the VPN with a non configurable forwarded port, because it's not possible to run Plex on anything else but 32400. Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

      - VPN_KEEP_LOCAL_DNS=false #
Setting this to true will re-add the default nameserver if it's been overwritten by the DNS = ... entry in wgO.conf. This should keep container name resolution working. If container name resolution still doesn't work, add --dns 1.1.1.1, somehow this can fix it (confirmed on MacOS).

      - VPN_FIREWALL_TYPE=auto #
Possible values are auto, legacy or nftables. The default is auto, this will try to use the most modern method available. If this doesn't work, you can try forcing it to legacy or nftables.

      - VPN_HEALTHCHECK_ENABLED=false
      - PRIVOXY_ENABLED=false
      - UNBOUND_ENABLED=false #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    cap_add:
      - NET_ADMIN
    sysctls:
      - net.ipv4.conf.all.src_valid_mark=1 #
Required in most cases, on some systems that don't have rp_filter set to strict, it's optional.

      - net.ipv6.conf.all.disable_ipv6=1 #
With net.ipv6.conf.all.disable_ipv6=1 all ipv6 support is disabled. Leave ipv6 disabled and remove all references to it in your wg0.conf file to keep things simple. If you need any sort of ipv6 support, enable it with net.ipv6.conf.all.disable_ipv6=0. A WireGuard ipv6 endpoint is currently not supported.

    devices:
      - /dev/net/tun:/dev/net/tun # OPTIONAL, READ THIS
If the WireGuard kernel module is missing (most likely on Synology/QNAP/Asustor), you can run WireGuard in userspace thanks to wireguard-go. For that you'll need to add the device /dev/net/tun. It's most likely that the device /dev/net/tun does not exist however, have a read here for instructions on checking and adding the device.

    ...

clicompose

docker run --rm \
    --hostname container-name.internal \ #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    -e VPN_ENABLED="true" \ #
There needs to be a file wg0.conf located in /config/wireguard and you need to set the variable VPN_ENABLED to true for the VPN to start. If you'd like to execute some of your own bash commands you can place two scripts alongside your wg0.conf, called wg0-pre.sh and wg0-post.sh. These will execute right before the check for the existence of a wg0.conf file and almost at the end, right before the internet connectivity test.

    -e VPN_CONF="wg0" \ # READ THIS
With VPN_CONF you can set the name used for your WireGuard config. This is an example of how your wg0.conf file should look like. If there's a lot of extra stuff, remove it unless you know what it's there for. The WireGuard config is automatically modified to use AllowedIPs = 0.0.0.0/1,128.0.0.0/1 for compatibility with Synology/QNAP/Asustor/WSL2 systems if you append -fix, so VPN_CONF=wg0-fix. WSL2 users can also compile their own kernel if they don't wanna use this workaround.
[Interface]
PrivateKey = supersecretprivatekey
Address = xx.xx.xxx.xxx/32 # Yes, /32
DNS = x.x.x.x

[Peer]
PublicKey = publickey
AllowedIPs = 0.0.0.0/0
Endpoint = xxx.x.xxx.x:51820


    -e VPN_PROVIDER="pia" \ #
Possible values are generic, proton and pia.

    -e VPN_LAN_NETWORK="192.168.1.0/24" \ #
The environment variable VPN_LAN_NETWORK can be set to for example 192.168.1.0/24, 192.168.1.0/24,192.168.44.0/24 or 192.168.1.33, so you can get access to the webui or other additional ports (see below). If for example you were to pick 192.168.0.0/24, every device with an ip in the range 192.168.0.0 - 192.168.0.255 on your LAN is allowed access to the webui.

    -e VPN_LAN_LEAK_ENABLED="false" \
    -e VPN_EXPOSE_PORTS_ON_LAN="" \ #
If you need to expose ports on your LAN you can use VPN_EXPOSE_PORTS_ON_LAN. For example VPN_EXPOSE_PORTS_ON_LAN=7878/tcp,9117/tcp, will block those ports on the vpn interface, so that there's no risk that they might be exposed to the world and allow access to them from your LAN. Some images also have a WEBUI_PORTS environment variable that does basically the same for the vpn part. For those apps that support it, it'll also change the port on which the app runs.

    -e VPN_AUTO_PORT_FORWARD="true" \ #
Auto retrieve a forwarded port and configure the supported app if set to true or if you can manually request/set a forwarded port in the VPN provider's web interface, fill in the port number (just the number). Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

    -e VPN_AUTO_PORT_FORWARD_TO_PORTS="" \ #
Adds a redirect for the forwarded port from your vpn provider to the internal port on which the app runs, ports in this list are also not blocked on the wireguard interface, so this var is also useful if you want to expose a port on both your LAN and VPN. Values like 32400/tcp will use the port from VPN_AUTO_PORT_FORWARD to create the redirect or if set to true the forwarded port from pia/proton. Use 3000@3001/tcp,3002@3003/tcp syntax for extra static redirects. The only known usecase as of right now is Plex and exposing it on the VPN with a non configurable forwarded port, because it's not possible to run Plex on anything else but 32400. Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

    -e VPN_KEEP_LOCAL_DNS="false" \ #
Setting this to true will re-add the default nameserver if it's been overwritten by the DNS = ... entry in wgO.conf. This should keep container name resolution working. If container name resolution still doesn't work, add --dns 1.1.1.1, somehow this can fix it (confirmed on MacOS).

    -e VPN_FIREWALL_TYPE="auto" \ #
Possible values are auto, legacy or nftables. The default is auto, this will try to use the most modern method available. If this doesn't work, you can try forcing it to legacy or nftables.

    -e VPN_HEALTHCHECK_ENABLED="false" \
    -e VPN_PIA_USER="" \ #
When using VPN_PROVIDER=pia, fill in your username and password. A wg0.conf will be automatically downloaded.

    -e VPN_PIA_PASS="" \
    -e VPN_PIA_PREFERRED_REGION="" \ #
By default a random server is used, but if you prefer a certain region you can fill in the region id. A list of available regions can be found in /config/wireguard after the first start. If you're seeing an error message shuf: getrandom: Function not implemented, you can't let it pick one randomly and are forced to fill in a region id.

    -e VPN_PIA_DIP_TOKEN="no" \ #
Fill in your DIP token here, if you've bought the dedicated ip option.

    -e VPN_PIA_PORT_FORWARD_PERSIST="false" \ #
If you'd like to keep using the same forwarded port until it expires, set this to true.

    -e PRIVOXY_ENABLED="false" \
    -e UNBOUND_ENABLED="false" \ #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    --cap-add=NET_ADMIN \
    --sysctl="net.ipv4.conf.all.src_valid_mark=1" \ #
Required in most cases, on some systems that don't have rp_filter set to strict, it's optional.

    --sysctl="net.ipv6.conf.all.disable_ipv6=1" \ #
With net.ipv6.conf.all.disable_ipv6=1 all ipv6 support is disabled. Leave ipv6 disabled and remove all references to it in your wg0.conf file to keep things simple. If you need any sort of ipv6 support, enable it with net.ipv6.conf.all.disable_ipv6=0. A WireGuard ipv6 endpoint is currently not supported.

    --device /dev/net/tun:/dev/net/tun \ # OPTIONAL, READ THIS
If the WireGuard kernel module is missing (most likely on Synology/QNAP/Asustor), you can run WireGuard in userspace thanks to wireguard-go. For that you'll need to add the device /dev/net/tun. It's most likely that the device /dev/net/tun does not exist however, have a read here for instructions on checking and adding the device.

    ...

services:
  app:
    hostname: container-name.internal #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    environment:
      - VPN_ENABLED=true #
There needs to be a file wg0.conf located in /config/wireguard and you need to set the variable VPN_ENABLED to true for the VPN to start. If you'd like to execute some of your own bash commands you can place two scripts alongside your wg0.conf, called wg0-pre.sh and wg0-post.sh. These will execute right before the check for the existence of a wg0.conf file and almost at the end, right before the internet connectivity test.

      - VPN_CONF=wg0 # READ THIS
With VPN_CONF you can set the name used for your WireGuard config. This is an example of how your wg0.conf file should look like. If there's a lot of extra stuff, remove it unless you know what it's there for. The WireGuard config is automatically modified to use AllowedIPs = 0.0.0.0/1,128.0.0.0/1 for compatibility with Synology/QNAP/Asustor/WSL2 systems if you append -fix, so VPN_CONF=wg0-fix. WSL2 users can also compile their own kernel if they don't wanna use this workaround.
[Interface]
PrivateKey = supersecretprivatekey
Address = xx.xx.xxx.xxx/32 # Yes, /32
DNS = x.x.x.x

[Peer]
PublicKey = publickey
AllowedIPs = 0.0.0.0/0
Endpoint = xxx.x.xxx.x:51820


      - VPN_PROVIDER=pia #
Possible values are generic, proton and pia.

      - VPN_LAN_NETWORK=192.168.1.0/24 #
The environment variable VPN_LAN_NETWORK can be set to for example 192.168.1.0/24, 192.168.1.0/24,192.168.44.0/24 or 192.168.1.33, so you can get access to the webui or other additional ports (see below). If for example you were to pick 192.168.0.0/24, every device with an ip in the range 192.168.0.0 - 192.168.0.255 on your LAN is allowed access to the webui.

      - VPN_LAN_LEAK_ENABLED=false
      - VPN_EXPOSE_PORTS_ON_LAN #
If you need to expose ports on your LAN you can use VPN_EXPOSE_PORTS_ON_LAN. For example VPN_EXPOSE_PORTS_ON_LAN=7878/tcp,9117/tcp, will block those ports on the vpn interface, so that there's no risk that they might be exposed to the world and allow access to them from your LAN. Some images also have a WEBUI_PORTS environment variable that does basically the same for the vpn part. For those apps that support it, it'll also change the port on which the app runs.

      - VPN_AUTO_PORT_FORWARD=true #
Auto retrieve a forwarded port and configure the supported app if set to true or if you can manually request/set a forwarded port in the VPN provider's web interface, fill in the port number (just the number). Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

      - VPN_AUTO_PORT_FORWARD_TO_PORTS= #
Adds a redirect for the forwarded port from your vpn provider to the internal port on which the app runs, ports in this list are also not blocked on the wireguard interface, so this var is also useful if you want to expose a port on both your LAN and VPN. Values like 32400/tcp will use the port from VPN_AUTO_PORT_FORWARD to create the redirect or if set to true the forwarded port from pia/proton. Use 3000@3001/tcp,3002@3003/tcp syntax for extra static redirects. The only known usecase as of right now is Plex and exposing it on the VPN with a non configurable forwarded port, because it's not possible to run Plex on anything else but 32400. Useful website to check for open ports is YouGetSignal and ipleak.net to leak test with .torrent file.

      - VPN_KEEP_LOCAL_DNS=false #
Setting this to true will re-add the default nameserver if it's been overwritten by the DNS = ... entry in wgO.conf. This should keep container name resolution working. If container name resolution still doesn't work, add --dns 1.1.1.1, somehow this can fix it (confirmed on MacOS).

      - VPN_FIREWALL_TYPE=auto #
Possible values are auto, legacy or nftables. The default is auto, this will try to use the most modern method available. If this doesn't work, you can try forcing it to legacy or nftables.

      - VPN_HEALTHCHECK_ENABLED=false
      - VPN_PIA_USER #
When using VPN_PROVIDER=pia, fill in your username and password. A wg0.conf will be automatically downloaded.

      - VPN_PIA_PASS
      - VPN_PIA_PREFERRED_REGION #
By default a random server is used, but if you prefer a certain region you can fill in the region id. A list of available regions can be found in /config/wireguard after the first start. If you're seeing an error message shuf: getrandom: Function not implemented, you can't let it pick one randomly and are forced to fill in a region id.

      - VPN_PIA_DIP_TOKEN=no #
Fill in your DIP token here, if you've bought the dedicated ip option.

      - VPN_PIA_PORT_FORWARD_PERSIST=false #
If you'd like to keep using the same forwarded port until it expires, set this to true.

      - PRIVOXY_ENABLED=false
      - UNBOUND_ENABLED=false #
When enabling the Unbound DNS server your requests will use DNS over TLS to Cloudflare. Except for requests made to .internal and .vpn TLDs, those are done to the local docker DNS server on 127.0.0.11. So if you want to use container hostnames to connect to other containers within a bridge network, you'll have to use --hostname and use container-name.internal or container-name.vpn. Currently .vpn is a non existing TLD, but that can change in the future. The TLD .internal should become the standard for internal networks, so it's the safest choice. Unbound can be used regardless of VPN_ENABLED being true or false.

    cap_add:
      - NET_ADMIN
    sysctls:
      - net.ipv4.conf.all.src_valid_mark=1 #
Required in most cases, on some systems that don't have rp_filter set to strict, it's optional.

      - net.ipv6.conf.all.disable_ipv6=1 #
With net.ipv6.conf.all.disable_ipv6=1 all ipv6 support is disabled. Leave ipv6 disabled and remove all references to it in your wg0.conf file to keep things simple. If you need any sort of ipv6 support, enable it with net.ipv6.conf.all.disable_ipv6=0. A WireGuard ipv6 endpoint is currently not supported.

    devices:
      - /dev/net/tun:/dev/net/tun # OPTIONAL, READ THIS
If the WireGuard kernel module is missing (most likely on Synology/QNAP/Asustor), you can run WireGuard in userspace thanks to wireguard-go. For that you'll need to add the device /dev/net/tun. It's most likely that the device /dev/net/tun does not exist however, have a read here for instructions on checking and adding the device.

    ...