The **CPU** requirements are pretty low, you can run it on a Raspberry Pi 1 A (ARMv6). The **RAM** required is around 100MiB (according to `docker stats`). The required disk **storage** depends on the content, please keep in mind, that the backend is a git repository that never forgets: Deleting a page or an attachment does not free up any space. The wiki needs no internet access. Clients using the wiki need only access to the server running the wiki, so it can run in an environment which is _isolated_ from the internet.

As URLs as dedicated domain (e.g. `wiki.domain.tld`) is required, it can not be mapped into a subfolder.

docker run --name otterwiki \

Open the wiki via http://127.0.0.1:8080 if you are running the docker command on your machine.

For all configuration options please see [[Configuration]].

1. Create a `docker-compose.yaml` file

2. Run `docker compose up -d`

3. Access the wiki via http://127.0.0.1:8080 if run on your machine.

4. Register your account. The first account is an admin-account with access to the application settings. The first accounts email address doesn't need to be confirmed nor has the account to be activated.

5. Customize the settings to your liking.

6. It's highly recommended to use a webserver as reverse proxy to connect to the docker process, see [Reverse Proxy](#reverse-proxy) below.

Alternativly you can configure the application using environment variables, for example:

For all configuration options please see [[Configuration]].

An Otter Wiki can be run with `podman` and `podman-compose` in the same way as with` docker` and `docker compose` please see above.

The image of An Otter Wiki runs as root user (with uid=0) and comes bundled with nginx that listens on port 80. The lighter image variant is the `-slim` image: It's based on alpine, has no nginx bundled and an uWSGI server is running as unprivileged `www-data` user (with uid=33). uWSGI serves both An Otter Wiki and the static files and listens on port 8080.

Except for the exposes port, the images are interchangeable. For example a `docker-compose.yaml` running the most recent version of a `-slim` image:

The `redimp/otterwiki:2-slim` image is as stable as the 'fat' image, after all it is what is running [otterwiki.com](https://otterwiki.com).

1. Install the prerequisites

ii. RHEL8 / Fedora / CentOS8 / Rocky Linux 8

2. Clone the otterwiki repository and enter the directory

3. Create and initialize the repository where the otterwiki data lives

mkdir -p app-data/repository

git init app-data/repository

4. Create a minimal `settings.cfg` e.g. via

5. Create a virtual environment and install An Otter Wiki

6. Run uwsgi listening on the localhost port 8080

7. Open http://127.0.0.1:8080 in your browser.

8. Register your account. The first account is an admin-account with access to the application settings.

9. Alternatively can you configure the application using the `settings.cfg`. For all configuration options please see [[Configuration]].

10. Create a service file e.g. `/etc/systemd/system/otterwiki.service`

It's highly recommended to use a webserver as reverse proxy to connect to uwsgi, see [Reverse Proxy](#reverse-proxy) below.

A reverse proxy is a server that sits in front of web servers and forwards

client (e.g. web browser) requests to those web servers. They are useful

when hosting multiple services on a host and make it much easier to configure

https. Neither An Otter Wiki itself nor the in the docker image provides https.

Mini how-tos for configuring Apache, NGINX and Caddy are provided below. For complete documention please check the corresponding software documention.

This is a minimal example of a config that configures NGINX as a reverse proxy. The full documentation about NGINX as reverse proxy can be found [here](https://docs.nginx.com/nginx/admin-guide/web-server/reverse-proxy/).

It's assumed that An Otter Wiki is running either in a docker container or as a uwsgi process and listening on port 8080.

- Install nginx via `apt install -y nginx`

- Create the `otterwiki.conf` in `/etc/nginx/sites-enabled/`

- Check the syntax via `nginx -t`

- Restart nginx via `systemctl restart nginx`

- Open <http://wiki.domain.tld> in your browser

- Check `journalctl -xeu nginx` and `/var/log/nginx/error.log` for errors.

- Install nginx via `dnf install nginx`

- Start NGINX and enable at boot: `sudo systemctl enable --now nginx`

- With SELinux enabled, make sure httpd can connect using `setsebool -P httpd_can_network_connect on`

- Create the `otterwiki.conf` in `/etc/nginx/conf.d/`

- Check the syntax via `nginx -t`

- Restart nginx via `systemctl restart nginx`

- Open <http://wiki.domain.tld> in your browser

- Check `journalctl -xeu nginx` and `/var/log/nginx/error.log` for errors.

See <https://www.redhat.com/sysadmin/setting-reverse-proxies-nginx> for a complete guide.

This is a minimal example how to configure Apache as reverse proxy. Please see the [Apache documentation](https://httpd.apache.org/docs/2.4/howto/reverse_proxy.html) for complete details.

It's assumed that An Otter Wiki is running either in a docker container or as a uwsgi process and listening on port 8080.

- Install apache2 via `apt install -y apache2`

- Enable proxy modules via `a2enmod proxy proxy_http`

- Create the `otterwiki.conf` config file in `/etc/apache2/site-available/`

- Enable the site via `a2ensite otterwiki`

- Restart apache2 `systemctl restart apache2`

- Open <http://wiki.domain.tld> in your browser

- Check `journalctl -xeu apache2.service` and `/var/log/apache2/error.log` for error messages.

- Install apache2 via `dnf install httpd`

- With SELinux enabled, make sure httpd can connect using `setsebool -P httpd_can_network_connect on`

- Restart apache2 via `systemctl restart httpd`

- Open <http://wiki.domain.tld> in your browser

- Check `journalctl -xeu httpd.service` and `/var/log/httpd/error_log` for error messages.

[Caddy](https://caddyserver.com/) is an open source webserver with a very light configuration that makes a fine reverse proxy.

After [installing](https://caddyserver.com/docs/install) caddy, configure `/etc/caddy/Caddyfile` with e.g.

With a server accessible from the internet, `domain.tld` beeing a proper domain name with A/AAAA DNS records pointing to the server, caddy will [automatically](https://caddyserver.com/docs/automatic-https) serve HTTPS.