Djeex/docudjeex
1
1
Fork 0
Code Issues Pull Requests Projects Wiki Activity

Migrate docudjeex to Docus v4 with EN/FR content

Browse Source
This commit is contained in:
Djeex
2026-08-30 16:41:37 +02:00
commit c51fcd5df6
241 changed files with 41143 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
content/en/3.serveex/3.security/.navigation.yml
+2
View File
@@ -0,0 +1,2 @@
title: Security
icon: i-lucide-shield
content/en/3.serveex/3.security/1.wireguard.md
+260
View File
@@ -0,0 +1,260 @@
---
title: Wireguard
description: Install and configure WireGuard VPN to securely access your homelab from anywhere and connect all your devices to your private network.
---
:ellipsis{left=0px width=40rem top=10rem blur=140px zIndex=60}
# Wireguard
::note
🎯 __Goals:__
- Install Wireguard
- Configure clients
- Access the secure network
::
## Introduction
---
Using a VPN allows remote access to a server’s local resources without exposing them to the internet. It’s a clean and secure way to access services like SSH without exposing the port publicly. With a VPN, you can securely connect to your network from anywhere and make devices on different networks communicate.
Here we will use [Wireguard](https://www.wireguard.com/), a secure and high-performance VPN server, using containers:
- [wg-easy](https://github.com/wg-easy/wg-easy) as the server, providing a very simple web UI to manage connections and download config files (including QR codes for phones)
- [Wireguard](https://docs.linuxserver.io/images/docker-wireguard/?h=wireguard) as the client for Linux systems
Clients are also available for Windows, macOS, iOS, and Android.
The concept:
- On the internet, anyone can reach any internet box and thus any exposed server.
- Your server is on your local network. It is accessible only locally unless services are explicitly exposed (as we did with Dockge). To access non-exposed resources, you must be on the same local network.
- We want to securely access these unexposed services (like SSH) from anywhere.
- We also want to connect services between servers, like linking two Dockge instances securely.
To achieve this, we’ll create a **Virtual Private Network** (VPN), i.e., a secure tunnel that only connected machines can use. They’ll appear to be on the same private network.
Additionally, you can add your phone, laptop, or other devices to the VPN and securely access your server resources wherever you are.
![picture](/img/serveex/vpn.svg)
In this diagram, machine 1 is part of two networks:
- Its local network (devices behind the same router, e.g. `192.168.x.x` – machines 1 and 2)
- The VPN network (VPN devices with a second IP, e.g. `10.8.x.x` – machines 1 and 4)
You *can* allow VPN clients to share access to their local networks, but we won’t do that here for security and subnet conflict reasons (e.g., if two remote machines use the same local IP like `192.168.1.1`).
So only VPN-connected devices can communicate with each other on the VPN, not with other local devices outside the VPN.
## Server Setup
---
::note
📋 **Pre-flight Checklist:**
- Ensure port `51820 UDP` is free on your server and correctly forwarded from your router (`51820 UDP -> Server`).
- Ensure port `51821 TCP` is free for the web UI.
::
::warning
__Warning__: If your IP is not static, use a Dynamic DNS service ([DynDNS](https://en.wikipedia.org/wiki/Dynamic_DNS)). If your ISP uses [CGNAT](https://en.wikipedia.org/wiki/Carrier-grade_NAT), you’ll need to use an external VPS and connect your local server as a client.
::
### Folder Structure
```sh
root
└── docker
└── wg-easy
├── config
│ └── etc_wireguard
├── compose.yaml
└── .env
```
Open Dockge, click **Compose**, and name the stack `wg_easy`.
Copy the following configuration:
```yaml
---
services:
wg-easy:
environment:
- INSECURE=true
image: ghcr.io/wg-easy/wg-easy:15
container_name: wg-easy
networks:
wg:
ipv4_address: 10.42.42.42
ipv6_address: fdcc:ad94:bacf:61a3::2a
volumes:
- ./etc_wireguard:/etc/wireguard
- /lib/modules:/lib/modules:ro
ports:
- "51820:51820/udp"
- "51821:51821/tcp"
restart: unless-stopped
cap_add:
- NET_ADMIN
- SYS_MODULE
sysctls:
- net.ipv4.ip_forward=1
- net.ipv4.conf.all.src_valid_mark=1
- net.ipv6.conf.all.disable_ipv6=0
- net.ipv6.conf.all.forwarding=1
- net.ipv6.conf.default.forwarding=1
networks:
wg:
driver: bridge
enable_ipv6: true
ipam:
driver: default
config:
- subnet: 10.42.42.0/24
- subnet: fdcc:ad94:bacf:61a3::/64
```
::tip
✨ **Tip:**
- You can customize WireGuard and web UI ports.
- Add a Watchtower label for automatic updates:
```yaml
services:
wg-easy:
# ...
labels:
- com.centurylinklabs.watchtower.enable=true
```
::
Deploy the stack and access the local web UI at `http://server-ip:51821`.
::caution
If the deployment fails, check your firewall rules.
::
Once connected, follow the web UI instructions to:
- Create your admin account and password.
- Set the host field (use your public IP or domain name).
Then go to *Administrator → Admin Panel → Config*:
- Change `Allowed IPs` from `0.0.0.0/24` to `10.8.0.0/24` for **split tunneling**.
- Remove IPv6 (it often causes unnecessary issues).
### Retrieve Configuration Files
To configure clients:
1. Access the web UI: `http://server-ip:51821`
2. Create a new client
3. Edit the client and add `10.8.0.0/24` to `Server Allowed IPs`
4. (Optional) Set `Persistent Keep Alive` to `25` if it’s a permanently connected client
5. Save, download, and rename the file to `wg0.conf` (or `wg1.conf`, etc.)
## Client Server Setup
---
::note
We assume the client server runs Linux with Docker installed.
::
### Folder Structure
```sh
root
└── docker
└── wireguard
└── config
│ └── wg_confs
└── compose.yaml
```
Create the folder:
```sh
sudo mkdir -p /docker/wireguard/config/wg_confs
```
::tip
✨ **Tip:** You can use [File Browser](/serveex/files/file-browser) instead of the terminal to edit and upload files.
::
Create the `wg0.conf` file:
```sh
sudo vi /docker/wireguard/config/wg_confs/wg0.conf
```
Enter insert mode (`i`), paste the downloaded configuration, then save (`Esc` → `:x`).
::tip
✨ **Alternative method:** Transfer the file via SFTP and move it:
```sh
sudo cp ~/wg0.conf /docker/wireguard/config/wg_confs
```
::
Create the `compose.yaml` file in `/docker/wireguard`:
```yaml
services:
wireguard:
image: lscr.io/linuxserver/wireguard:latest
container_name: wireguard
network_mode: host
cap_add:
- NET_ADMIN
- SYS_MODULE
environment:
- TZ=Europe/Paris
volumes:
- /docker/wireguard/config:/config
- /lib/modules:/lib/modules
restart: unless-stopped
```
Start the container:
```sh
cd /docker/wireguard
sudo docker compose up -d
```
::note
Repeat this setup for each client.
::
## Other Devices
---
- **Mobile:** Install WireGuard and scan the QR code via the web UI (`http://server-ip:51821`)
- **Desktop:** Install the WireGuard client and import the downloaded config file.
::warning
**Note:** If the client machine is on the same local network as the server, edit the `wg0.conf` file to use the local server IP:
`Endpoint = server-local-ip:51820`
::
And here’s the final setup overview:
![picture](/img/serveex/wireguard.svg)
content/en/3.serveex/3.security/2.authentik.md
+588
View File
@@ -0,0 +1,588 @@
---
title: Authentik
description: Install Authentik as a self-hosted identity provider — configure MFA and protect your services with SSO and reverse proxy authentication.
---
:ellipsis{left=0px width=40rem top=10rem blur=140px zIndex=60}
# Authentik
::note
🎯 __Objectives:__
- Install and expose Authentik
- Configure Multi-Factor Authentication (MFA)
- Protect a native app or an app behind a reverse proxy
::
[Authentik](https://goauthentik.io) is a single sign-on (SSO) tool that allows you to log in once to all platforms compatible with OpenID. It can also secure access to your exposed services by injecting itself via SWAG into requests to those services.
For example, if you're exposing Dockge online at `dockge.mydomain.com`, you’ll first land on an Authentik login page when accessing it. If you've already authenticated with another Authentik-protected service, you won’t need to log in again. This allows you to authenticate only once per day for all protected services.
Authentik also supports multi-factor authentication, including TOTP (a code generated by the authentication app of your choice). Additionally, it allows login through Microsoft or Google accounts, provided you've configured one of those applications.
It's a great alternative to VPNs for securely exposing services, especially ones that lack MFA or login protection (e.g., the SWAG dashboard).
Authentik has [extensive documentation](https://docs.goauthentik.io/docs/installation/docker-compose) and [great tutorials from Cooptonian](https://www.youtube.com/@cooptonian). Here, we’ll cover the basics using Dockge as an example.
There are two main modes you should know:
- The first allows apps with native support for OpenID-compatible SSO to connect directly to Authentik. This is the preferred method, as the app itself decides what’s public and what’s protected.
![Picture](/img/serveex/auth-native.svg)
- The second method injects Authentik authentication through SWAG before reaching the target service.
![Picture](/img/serveex/auth-proxy.svg)
Both modes can be configured on a per-application basis.
## Installation
---
Folder structure:
```sh
root
└── docker
└── authentik
├── .env
├── compose.yml
├── media
├── certs
├── custom-template
└── ssh
```
Create the folders:
```sh
sudo mkdir -p /docker/authentik/media /docker/authentik/certs /docker/authentik/custom-template /docker/authentik/ssh
```
Navigate to the `authentik` folder via `cd /docker/authentik` and generate a password and secret key to include in the `.env` file:
```sh
sudo echo "PG_PASS=$(openssl rand 36 | base64)" >> .env
sudo echo "AUTHENTIK_SECRET_KEY=$(openssl rand 60 | base64)" >> .env
```
::note
To generate the keys, we created the folders ahead of deployment using Dockge. Dockge will prevent you from creating a stack with the same name in these folders unless a `compose.yml` file exists. So, create an empty `compose.yml` so it appears as an inactive stack:
```sh
sudo vi /docker/authentik/compose.yml
```
::
Open Dockge and search for "authentik" in the inactive stacks.
Name the stack `authentik` and paste the following configuration, replacing `{AUTHENTIK_TAG:-2026.2}`{lang=properties} with [the latest version of Authentik](https://goauthentik.io/docs/releases).
```yaml
---
services:
postgresql:
image: docker.io/library/postgres:16-alpine
container_name: authentik-postgresql
restart: unless-stopped
healthcheck:
test:
- CMD-SHELL
- pg_isready -d $${POSTGRES_DB} -U $${POSTGRES_USER}
start_period: 20s
interval: 30s
retries: 5
timeout: 5s
volumes:
- database:/var/lib/postgresql/data
environment:
POSTGRES_PASSWORD: ${PG_PASS:?database password required}
POSTGRES_USER: ${PG_USER:-authentik}
POSTGRES_DB: ${PG_DB:-authentik}
env_file:
- .env
redis:
image: docker.io/library/redis:alpine
container_name: authentik-redis
command: --save 60 1 --loglevel warning
restart: unless-stopped
healthcheck:
test:
- CMD-SHELL
- redis-cli ping | grep PONG
start_period: 20s
interval: 30s
retries: 5
timeout: 3s
volumes:
- redis:/data
server:
image: ${AUTHENTIK_IMAGE:-ghcr.io/goauthentik/server}:${AUTHENTIK_TAG:-2026.2}
container_name: authentik-server
restart: unless-stopped
command: server
environment:
AUTHENTIK_REDIS__HOST: redis
AUTHENTIK_POSTGRESQL__HOST: postgresql
AUTHENTIK_POSTGRESQL__USER: ${PG_USER:-authentik}
AUTHENTIK_POSTGRESQL__NAME: ${PG_DB:-authentik}
AUTHENTIK_POSTGRESQL__PASSWORD: ${PG_PASS}
volumes:
- ./media:/media
- ./custom-templates:/templates
- ./ssh:/authentik/.ssh
env_file:
- .env
ports:
- ${COMPOSE_PORT_HTTP:-9000}:9000
- ${COMPOSE_PORT_HTTPS:-9443}:9443
depends_on:
- postgresql
- redis
worker:
image: ${AUTHENTIK_IMAGE:-ghcr.io/goauthentik/server}:${AUTHENTIK_TAG:-2026.2}
container_name: authentik-worker
restart: unless-stopped
command: worker
environment:
AUTHENTIK_REDIS__HOST: redis
AUTHENTIK_POSTGRESQL__HOST: postgresql
AUTHENTIK_POSTGRESQL__USER: ${PG_USER:-authentik}
AUTHENTIK_POSTGRESQL__NAME: ${PG_DB:-authentik}
AUTHENTIK_POSTGRESQL__PASSWORD: ${PG_PASS}
# `user: root` and the docker socket volume are optional.
# See more for the docker socket integration here:
# https://goauthentik.io/docs/outposts/integrations/docker
# Removing `user: root` also prevents the worker from fixing the permissions
# on the mounted folders, so when removing this make sure the folders have the correct UID/GID
# (1000:1000 by default)
user: root
volumes:
- /var/run/docker.sock:/var/run/docker.sock
- ./media:/media
- ./certs:/certs
- ./custom-templates:/templates
- ./ssh:/authentik/.ssh
env_file:
- .env
depends_on:
- postgresql
- redis
volumes:
database:
driver: local
redis:
driver: local
```
In the `.env` file, the `PG_PASS` and `AUTHENTIK_SECRET_KEY` variables are already set.
Deploy the stack.
You can then begin the initial setup by visiting:
`http://yourserverip:9000/if/flow/initial-setup/`
::warning
__Warning:__ It’s recommended to create a new admin account and **disable** the default `akadmin` account.
::
## Exposing Authentik
---
To use Authentik outside your local network, you must expose it.
::note
📋 __Prerequisites:__ <br/><br/>
We assume you have already created a subdomain like `auth.mydomain.com` in your [DNS zone](/general/networking/dns), with a CNAME pointing to `mydomain.com`. Also, unless you're using [Cloudflare Zero Trust](/serveex/security/cloudflare), you must have already forwarded port `443` from your router to port `443` of your server in your [NAT rules](/general/networking/nat).
::
Open the `authentik-server.conf` file:
::tip
✨ __Tip for those who dislike terminals:__
You can use [File Browser](/serveex/files/file-browser) to navigate and edit files instead of using terminal commands.
::
```sh
sudo vi /docker/swag/config/nginx/authentik-server.conf
```
Verify that the following variables are set correctly:
```nginx
set $upstream_authentik authentik-server;
proxy_pass http://$upstream_authentik:9000;
```
If not, press `i` to enter edit mode, make the necessary changes, then save and exit by pressing `Esc` followed by `:x`.
Create the `auth.subdomain.conf` file:
```sh
sudo vi /docker/swag/config/nginx/proxy-confs/auth.subdomain.conf
```
Press `i` to enter edit mode and paste the following configuration:
```nginx
## Version 2023/05/31
# Ensure your authentik container is named authentik-server
# Ensure your DNS has a CNAME for authentik
server {
listen 443 ssl;
listen [::]:443 ssl;
server_name auth.*;
include /config/nginx/ssl.conf;
client_max_body_size 0;
location / {
include /config/nginx/proxy.conf;
include /config/nginx/resolver.conf;
set $upstream_app authentik-server;
set $upstream_port 9000;
set $upstream_proto http;
proxy_pass $upstream_proto://$upstream_app:$upstream_port;
}
location ~ (/authentik)?/api {
include /config/nginx/proxy.conf;
include /config/nginx/resolver.conf;
set $upstream_app authentik-server;
set $upstream_port 9000;
set $upstream_proto http;
proxy_pass $upstream_proto://$upstream_app:$upstream_port;
}
}
```
Save and exit by pressing `Esc` then `:x`.
Go to Dockge, and edit the SWAG compose file to add the Authentik network:
```yaml
services:
swag:
container_name: # ...
# ...
networks: # Attach the container to the custom network
# ...
- authentik # Name of the network declared in the stack
networks: # Define the custom network
# ...
authentik: # Name of the network declared in the stack
name: authentik_default # Actual name of the external network
external: true # Indicates it's an external network
```
Restart the stack and wait for SWAG to be fully operational.
Done! You can now access Authentik via `https://auth.mydomain.com`
## Enable Multifactor Authentication
---
The main value of Authentik is using multifactor authentication for all protected apps.
- Go to `https://auth.mydomain.com`
- Log in
- Go to _Settings_
- Click the _MFA_ section
- Click _Register_
- Choose a method like _TOTP device_ (you'll need an authenticator app like Google Authenticator)
- Follow the steps
You’ll now be prompted to enter a one-time code at every login.
## Protecting a Native App
---
Authentik is natively compatible with several applications. You can find the list and [support here](https://docs.goauthentik.io/integrations/services/).
## Protecting an App via Reverse Proxy
---
SWAG lets you insert Authentik’s login page between a request and access to your service. To do this:
- Configure the authentication provider in Authentik.
- Edit the domain proxy file so SWAG can intercept the request.
Why do this when Dockge already has authentication? Because Dockge uses weak HTTP authentication. With Authentik, you get strong MFA authentication and automatic login to all apps protected by Authentik. This secures access to Dockge and other apps without needing a VPN.
### Configuring Authentik
- Go to Authentik
- Open the admin panel
- Select _Applications_ then _Create with wizard_
- Fill in the fields as shown:
![Picture](/img/serveex/auth1.png)
- At the next step, choose "Forward authentication (single application)" and configure it as shown (flows are important):
![Picture](/img/serveex/auth2.png)
- Next, go to the _Outposts_ menu on the left and edit _authentik Embedded Outpost_:
![Picture](/img/serveex/auth3.png)
- Add the `dockge` application by moving it to the right column and save.
### Configuring SWAG
Edit the file `dockge.mydomain.com`:
```sh
sudo vi /docker/swag/config/nginx/proxy-confs/dockge.subdomain.conf
```
Press `i` to enter edit mode and uncomment the two lines `#include /config/nginx/authentik-server.conf;`
Press `Esc`, type `:x`, and press `Enter` to save and exit.
Done! Now when accessing `https://dockge.mydomain.com`, you’ll be redirected to the Authentik login screen.
::tip
✨ __Tip:__ In Dockge's settings, you can disable Dockge's authentication to avoid double login. **Warning**: this means if the port is open on your local network, there will be no authentication at all.
::
::note
Repeat this process for each app you want to protect (unless it has native integration with Authentik).
::
Your new architecture looks like this:
![Picture](/img/serveex/authentik.svg)
## Protecting a Remote Server Service
---
For a [native application](/serveex/security/authentik/#protecting-a-native-app) (via OAuth 2.0 or other), nothing changes.
For a non-native app behind a reverse proxy, you must deploy an __Outpost__. An Outpost is a container acting as a local proxy — it's the target of your app's auth requests and the only one authorized to communicate with your Authentik API.
::note
Prerequisites:
- Install [Docker](/serveex/core/docker) on the remote server hosting the service.
- If the app has no native integration, use a compatible reverse proxy. We will use [SWAG](/serveex/core/swag) here.
::
This container will forward requests to your main [Authentik](/serveex/security/authentik#authentik) instance over the internet (or your local network). The server will perform checks and respond to the Outpost, which will allow or block access accordingly.
![auth-outpost](/img/serveex/auth-outpost.svg)
### Configuring Authentik
Create your [providers and applications](/serveex/security/authentik/#protecting-a-native-app) as shown earlier.
Then, in the admin panel, go to _Applications > Outposts_, and create a new outpost.
Fill in as follows:
| Field | Value |
|----------------|------------------------------------------------------------------------|
| `Name` | Your preferred name |
| `Type` | `Proxy` |
| `Integration` | Leave empty |
| `Applications` | Select the applications you previously created |
In the `Advanced settings` section, clear the existing content and enter:
```yaml
log_level: info
docker_labels: null
authentik_host: https://your_authentik_server_domain/
object_naming_template: ak-outpost-%(name)s
authentik_host_insecure: false
container_image:
docker_network: null
docker_map_ports: true
docker_labels: null
```
Save and exit.
On the list of created outposts, locate the new one and click _Show details_ at the end of the line. Carefully copy the access token.
### Configuring the Remote Machine
We assume you’ve already installed [Docker](/serveex/core/docker) and [SWAG](/serveex/core/swag) on this remote machine.
On your remote machine, use [Dockge](/serveex/core/docker/#installer-dockge-pour-gérer-et-déployer-les-conteneurs) to create a stack named `authentik-outpost`.
If you haven’t installed [Dockge](/serveex/core/docker/#installer-dockge-pour-gérer-et-déployer-les-conteneurs), create a folder `/docker/authentik-outpost`, or directly via command line:
```sh
sudo mkdir -P /docker/authentik-outpost
```
::tip
✨ __Tip for terminal-averse users:__
You can use [File Browser](/serveex/files/file-browser) to navigate and edit your files instead of using terminal commands.
::
Create the `compose.yaml` file or paste the configuration directly into Dockge if installed.
Via command line:
```sh
sudo vi /docker/authentik-outpost/compose.yaml
```
Enter edit mode by pressing `i` and paste the following configuration, updating the version in `{AUTHENTIK_TAG:proxy:2024.2.3}`{lang=properties} to match your Authentik server version.
```yaml
version: "3.5"
services:
authentik_proxy:
container_name: authentik-outpost
image: ghcr.io/goauthentik/proxy:2024.2.3
# Optionally specify which networks the container should be
# might be needed to reach the core authentik server
restart: unless-stopped
env_file:
- .env
ports:
- 9000:9000
- 9443:9443
environment:
AUTHENTIK_HOST: ${HOST}
AUTHENTIK_INSECURE: "false"
AUTHENTIK_TOKEN: ${TOKEN}
```
Go to the SWAG stack on the remote machine (or edit directly using Dockge) and add the authentik-outpost network in the configuration file like this (see `networks` section):
```sh
sudo vi /docker/swag/compose.yaml
```
```yaml
services:
swag:
container_name: #...
# ...
networks: # Attach the container to the custom network
- authentik-outpost # Network name as declared in the stack
networks: # Define the custom network
#...
authentik-outpost: # Name of the network declared in the stack
name: authentik-outpost_default # Actual name of the external network
external: true # Marks it as an external network
```
Press `Esc`, then type `:x` and press `Enter` to save and exit.
::note
We assume the Dockge network name is `authentik-outpost_default`.
::
If using [Dockge](/serveex/core/docker/#installer-dockge-pour-gérer-et-déployer-les-conteneurs), restart SWAG.
Otherwise, via terminal:
```sh
cd /docker/swag/
sudo docker compose up -d
```
Create (or fill using Dockge) the `.env` file in the `authentik-outpost` directory:
Via command line:
```sh
sudo vi /docker/authentik-outpost/.env
```
Enter edit mode with `i` and paste the following configuration:
```properties
HOST=
TOKEN=
```
Fill in the values:
| Variable | Value | Example |
|----------|-------|---------|
| `HOST`{lang=properties} | The URL of your Authentik server | `https://auth.domain.com` |
| `TOKEN`{lang=properties} | The previously copied access token | `Q2pVEqsTNRkJSO9SkJzU3KZ2` |
Press `Esc`, then type `:x` and press `Enter` to save and exit.
If using Dockge, deploy the stack.
Otherwise, via terminal:
```sh
cd /docker/authentik-outpost/
sudo docker compose up -d
```
The container is now running. You can verify its status from your Authentik instance admin panel under _Applications > Outposts_.
Now, let’s configure SWAG.
Open the `authentik-server.conf` file:
```sh
sudo vi /docker/swag/config/nginx/authentik-server.conf
```
In the file, press `i` to enter edit mode and change `authentik-server` to `authentik-outpost` as shown:
```nginx
set $upstream_authentik authentik-outpost;
proxy_pass http://$upstream_authentik:9000;
```
Save and exit with `Esc`, then `:x` and `Enter`.
Then configure the applications to protect as you did on your main server, whether they are [native](/serveex/security/authentik/#protecting-a-native-app) or protected via [reverse proxy](/serveex/security/authentik#protecting-an-app-via-reverse-proxy).
## Migrating an Authentik Database
---
On the source machine, dump the database:
```sh
sudo docker exec authentik-postgres pg_dump -U authentik -F t authentik > /path/to/mydb.tar
```
Then transfer it to the target machine. On the target machine, copy the file into the Docker container:
```sh
cp /path/to/mydb.tar authentik-postgres:/path/to/wherever
```
(Optional) Purge existing tables:
```sh
sudo docker exec -i authentik-postgres psql -U authentik -c "SELECT pg_terminate_backend(pg_stat_activity.pid) FROM pg_stat_activity WHERE pg_stat_activity.datname = 'authentik' AND pid <> pg_backend_pid();" && sudo docker exec -i authentik-postgres psql -U authentik -d postgres -c "DROP DATABASE IF EXISTS authentik;" && sudo docker exec -i authentik-postgres psql -U authentik -d postgres -c "CREATE DATABASE authentik;"
```
Restore the database:
```sh
sudo docker exec authentik-postgresql pg_restore -U authentik -d authentik /path/to/wherever/mydb.tar
```
content/en/3.serveex/3.security/3.cloudflare.md
+272
View File
@@ -0,0 +1,272 @@
---
title: Cloudflare Zero Trust
description: Use Cloudflare Tunnels and Zero Trust to expose homelab services without opening ports — configure SWAG and manage multiple tunnels.
---
:ellipsis{left=0px width=40rem top=10rem blur=140px zIndex=60}
# Cloudflare Zero Trust
::note
🎯 __Goals:__
- Understand the concept of Cloudflare Tunnels
- Configure your Cloudflare account
- Configure SWAG
- Manage multiple tunnels
::
![cloudfare_tunnels](/img/serveex/cloudflared.svg)
## Introduction
---
The _Zero Trust_ architecture is the practice of designing systems based on the principle of __"never trust, always verify"__, as opposed to the traditional principle of __"trust, but verify"__. This concept has become increasingly popular recently due to the growing number of attacks targeting user data. It’s a broad concept, but we’ll focus on how to apply _Zero Trust_ to the web services we host.
_Cloudflare tunnels_ offer a simple way to implement _Zero Trust_, using [SWAG](/serveex/core/swag) and [Authentik](/serveex/security/authentik).
Simply put, Cloudflare Tunnels allow you to:
- Hide your server’s IP (and your home IP if it's self-hosted)
- Authenticate traffic
- Benefit from Cloudflare protections (DDoS attacks, blacklists, malicious requests, etc.)
- Use Cloudflare's CDN to cache and speed up your websites
- Avoid opening router ports for services exposed by SWAG
Here we’ll explain how to integrate SWAG with Cloudflare tunnels.
::warning
- __Warning:__
- Do not use Cloudflare tunnels to expose a mail server
- Do not use Cloudflare tunnels to expose a video service like Plex (if you followed [this guide](/serveex/media/plex), Plex is not exposed, so it’s fine)
- Do not use Cloudflare tunnels for the BitTorrent protocol (if you followed [this guide](/serveex/media/qbittorrent), everything is fine)
::
## Cloudflare Configuration
---
### DNS Zone
First, you need to set Cloudflare as your [DNS zone](/general/networking/dns) manager. If you bought your domain from Cloudflare, that’s already done. Otherwise, check with your registrar how to add external DNS servers. Cloudflare provides [step-by-step documentation](https://developers.cloudflare.com/dns/zone-setups/full-setup/setup/) on how to configure a DNS Zone, whether your domain is external or registered with Cloudflare.
If you only have one server to protect behind Cloudflare, you can delete all existing DNS records. By default, your domain and all its subdomains will be redirected to the tunnel.
If you have subdomains pointing to other servers, you can still define them in the DNS zone using A records.
If you have several servers and tunnels under one domain, [see here](http://192.168.7.80:8005/serveex/cloudflare/#gerer-plusieurs-tunnels-pour-plusieurs-serveurs).
### API Key
Start by creating a new Cloudflare API token and retrieving your zone and account IDs.
On your Cloudflare dashboard, on your domain overview page, you’ll see the `zone` and `account` IDs at the bottom right. Save both securely.
![id and account](/img/serveex/cf-id.png)
Just below that is a link titled _Get your API token_. Click it. The token scope must include `Zone:DNS:Edit` and `Account:Cloudflare Tunnel:Edit`. Your page should look like the screenshot below.
![API token](/img/serveex/cf-token.png)
Once created, your token will only be shown once. Save it securely, as it cannot be viewed again later.
### Cloudflare Zero Trust
You must register for _Cloudflare Teams_ to access the _Zero Trust_ dashboard that manages tunnels and access policies. This is a premium service, but there’s a free plan for up to 50 users—perfect for a home lab. Keep in mind that a valid credit card is required to register, but the free plan incurs no charges.
Register [via this link](https://dash.teams.cloudflare.com/).
## SWAG Configuration
---
::note
This guide assumes you own `mondomaine.fr` and that its DNS is correctly pointing to Cloudflare, as described above.
::
SWAG supports two Docker Mods:
- __Cloudflared__, the container used to create and manage tunnels
- __Cloudflared Real IP__, which allows SWAG to receive the true source IP of incoming requests instead of Docker’s internal IP (important for IP geolocation mods like DBIP).
These two mods, merged into the SWAG container, require some configuration.
### Tunnel Configuration
Create a file `tunnelconfig.yml` to reference in your SWAG `compose.yaml`.
::tip
✨ __Tip:__ Use [File Browser](/serveex/files/file-browser) to navigate and edit files instead of using the terminal.
::
```sh
sudo vi /docker/swag/config/tunnelconfig.yml
```
Press `i` to enter insert mode and paste:
```yaml
ingress:
- hostname: mondomaine.fr
service: https://mondomaine.fr
- hostname: "*.mondomaine.fr"
service: https://mondomaine.fr
- service: http_status:404
```
Press `Esc`, then save and exit with `:x` and `Enter`.
### Cloudflare Real IP Configuration
Now configure _Cloudflare Real IP_.
Open the `nginx.conf` file:
```sh
sudo vi /docker/swag/config/nginx/nginx.conf
```
Press `i` and add the following at the end of the `http` section:
```nginx
real_ip_header X-Forwarded-For;
real_ip_recursive on;
include /config/nginx/cf_real-ip.conf;
set_real_ip_from 127.0.0.1;
```
Save and exit with `:x`.
### Docker Compose
In Dockge, edit your SWAG stack with this:
```yaml
---
services:
swag:
image: lscr.io/linuxserver/swag:latest
container_name: swag
cap_add:
- NET_ADMIN
env_file:
- .env
environment:
- DOCKER_MODS=linuxserver/mods:swag-dbip|linuxserver/mods:swag-dashboard|linuxserver/mods:swag-auto-reload|linuxserver/mods:universal-cloudflared|linuxserver/mods:swag-cloudflare-real-ip
- PUID=${PUID}
- PGID=${PGID}
- TZ=Europe/Paris
- URL=${DOMAIN}
- SUBDOMAINS=wildcard
- VALIDATION=dns
- DNSPLUGIN=${PLUGIN}
- EMAIL=${EMAIL}
- CF_ZONE_ID=${ZONE_ID}
- CF_ACCOUNT_ID=${ACCOUNT_ID}
- CF_API_TOKEN=${API_TOKEN}
- CF_TUNNEL_NAME=${TUNNEL_NAME}
- CF_TUNNEL_PASSWORD=${TUNNEL_PW}
- FILE__CF_TUNNEL_CONFIG=/config/tunnelconfig.yml
extra_hosts:
- ${DOMAIN}:127.0.0.1
ports:
- 81:81
volumes:
- /docker/swag/config:/config
- /docker/swag/config/fail2ban/fail2ban.sqlite3:/dashboard/fail2ban.sqlite3:ro
restart: unless-stopped
```
::tip
✨ __Tip:__ Add a Watchtower label to automate updates:
```yaml
labels:
- com.centurylinklabs.watchtower.enable=true
```
::
Fill in your `.env` file:
```properties
PUID=
PGID=
DOMAIN=
PLUGIN=
EMAIL=
ZONE_ID=
ACCOUNT_ID=
API_TOKEN=
TUNNEL_NAME=
TUNNEL_PW=
```
| Variable | Value | Example |
|----------------|-------------------------------------------------------------|--------------------------------|
| `PUID` | User ID (`id username`) | `1000` |
| `GUID` | Group ID (`id username`) | `1000` |
| `DOMAIN` | Your reserved domain | `mondomaine.fr` |
| `PLUGIN` | DNS provider (also configure `cloudflare.ini`) | `cloudflare` |
| `EMAIL` | Email for the certificate | `[email protected]` |
| `ZONE_ID` | Cloudflare Zone ID | `aNhcz1l3JfWbFZo2XMpzQlP2iOqk` |
| `ACCOUNT_ID` | Cloudflare Account ID | `buKsjNHLyzKMM1qYnzOy4s7SHfly` |
| `API_TOKEN` | API token | `53ydYus9TFFk1DOXNdP87iIcJtQjoW` |
| `TUNNEL_NAME` | Tunnel name | `my_tunnel` |
| `TUNNEL_PW` | Strong, random password | `iSzKRmP4VbnlsMvdSdgBEJiJi` |
Once done, deploy the stack. Check the logs—you should reach `server ready`.
Then confirm your tunnel appears under _Networks > Tunnels_ in [Cloudflare Zero Trust](https://one.dash.cloudflare.com/). By default, all subdomains will be routed through the tunnel—no need to define them [in your DNS zone](/general/networking/dns).
::tip
✨ __Tip:__ If you want to expose a service without a tunnel, just define an A record [in your DNS zone](/general/networking/dns). If resolution fails, disable the proxy function for that record—e.g., for `sub.mondomaine.fr`.
![dns](/img/serveex/cf-dns.png)
::
## Managing Multiple Tunnels for Multiple Servers
---
By default, all subdomains of your domain are routed through the single tunnel. But if you have a second server, just change the tunnel name in that SWAG instance.
In your DNS zone, redirect subdomains to the correct tunnel.
Go to _Networks > Tunnels_ in [Cloudflare Zero Trust](https://one.dash.cloudflare.com/).
Note the tunnel IDs:
![tunnels_id](/img/serveex/cf-tunnels-id.png)
Then in the [Cloudflare DNS dashboard](https://dash.cloudflare.com/), click your domain name.
Click `Add Record` and add these two CNAME records (include `.cfargotunnel.com`):
| Type | Name | Target |
|---------|--------------|----------------------------------------|
| `CNAME` | `subdomain1` | `yourtunnelid1.cfargotunnel.com` |
| `CNAME` | `subdomain2` | `yourtunnelid2.cfargotunnel.com` |
If you have many subdomains, point them to the above reference subdomains.
This way, if a tunnel ID changes, you only update one DNS record.
Example:
- `sub1` and `sub2` also point to the server behind `subdomain1`:
| Type | Name | Target |
|---------|--------|---------------|
| `CNAME` | `sub1` | `subdomain1` |
| `CNAME` | `sub2` | `subdomain1` |
- `sub3` and `sub4` point to the server behind `subdomain2`:
| Type | Name | Target |
|---------|--------|---------------|
| `CNAME` | `sub3` | `subdomain2` |
| `CNAME` | `sub4` | `subdomain2` |