WARPStack

https://gitlab.com/Misaka-blog/warp-script
Gluetun
https://github.com/qdm12/gluetun-wiki
In the download server, stored in ~/docker/warpstack and in individual folders megabasterd and gluetun

Docker-compose

services:
  gluetun:
    image: qmcgaw/gluetun
    container_name: gluetun
    cap_add:
      - NET_ADMIN
    devices:
      - /dev/net/tun:/dev/net/tun
    ports:
      - 8888:8888/tcp # HTTP proxy
      - 8388:8388/tcp # Shadowsocks
      - 8388:8388/udp # Shadowsocks
      - 5861:5860
    volumes:
      - ~/docker/gluetun:/gluetun
    env_file:
      - .env # keys
      - wg.env # ip and port
    environment:
      - VPN_SERVICE_PROVIDER=custom
      - VPN_TYPE=wireguard
      - WIREGUARD_ADDRESSES=172.16.0.2/32
      - FIREWALL_OUTBOUND_SUBNETS=10.10.120.0/24

Additional configuration such as NET_ADMIN and the /dev/net/tun are needed.
The ports 5861:5860 is for forwarding ports of other containers

• the port 5861 is the port accessible on the host and bind to 5860 in the container
• the port 5860 is added via other container

The environments are split into 3 parts

• environments: non-secret general configurations
• .env Wireguard public and private keys which should change
• wg.env Wireguard IP address and port information, it is separate to allow easy changing by other programs

FIREWALL_OUTBOUND_SUBNETS - this is the subnet for local network outside of Docker, it will be used so the container under VPN will also be able to connect to other computers outside Docker

Adding containers to Wireguard VPN

  network_mode: "service:gluetun" # container:gluetun if outside of compose stack

Uses special network mode, no port mapping is needed as it won’t work. The port has to be set on application level.
Eg. usually 81:80 is required for other computers to access web port on a container, but when it’s under gluetun,. The containers port 80 is mapped to gluetun’s port 80 and port forwarding needs to be done via gluetun. If multiple containers runs on port 80 internally, the port must be changed via application settings.

MegaBasterd

megabasterd:
    image: vttc08/megabasterd
    network_mode: "service:gluetun"
    volumes:
      - "~/docker/warpstack/megabasterd:/config:rw"
      - "/mnt/nvme/share/VNC/megabasterd_dl:/output:rw"
    environment:
      - VERSION=8.14
      - KEEP_APP_RUNNING=1
      - WEB_LISTENING_PORT=5860
    env_file: ~/docker/vnc/vnc.env
    restart: unless-stopped
    container_name: megabasterd

Use the environment WEB_LISTENING_PORT=5860 to change the port in case multiple jlesage containers are needed behind a VPN, since all jlesage containers runs on port 5800.

ADD SQL OPTIONS LATER

run_command_path|/config/reset.sh

kill -s SIGTERM 1 # kill a container from inside

Change this in MegaBasterd, by default jlesage apps including this have configuration files located in /config, create reset.sh with the following content.

#!/bin/sh
wget --post-data='{"actionId": "reset-megabasterd"}' --header='Content-Type: application/json' http://10.10.120.16:1378/api/StartAction

• 10.10.120.16:1378 is the OliveTin server
• uses wget instead of curl to POST data as it’s the default included in jlesage containers
• the OliveTin has action ID of reset-megabasterd
• uses OliveTin to execute scripts which restart the entire stack