Nextcloud with podman rootless containers and user systemd services. Part IV - Exposing Nextcloud externally

Jan 28, 2021 08:30 · 783 words · 4 minute read

Introducing bunkerized-nginx 🔗

I heard about bunkerized-nginx a while ago and I thought it would be nice to use it as a reverse proxy so I can expose my internal services to the internet ‘safely’.

A non-exhaustive list of features (copy & paste from the README):

  • HTTPS support with transparent Let’s Encrypt automation
  • State-of-the-art web security : HTTP security headers, prevent leaks, TLS hardening, …
  • Integrated ModSecurity WAF with the OWASP Core Rule Set
  • Automatic ban of strange behaviors with fail2ban
  • Antibot challenge through cookie, javascript, captcha or recaptcha v3
  • Block TOR, proxies, bad user-agents, countries, …
  • Block known bad IP with DNSBL and CrowdSec
  • Prevent bruteforce attacks with rate limiting
  • Detect bad files with ClamAV
  • Easy to configure with environment variables or web UI
  • Automatic configuration with container labels

A must have for me was having support for Let’s Encrypt and having an easy way to configure it. So this was a perfect match to me!

Firewall ports 🔗

As the container is going to be rootless, we need to open a few ports in the host as root. We will use 8080/tcp and 8443/tcp:

sudo -s -- sh -c \
  "firewall-cmd -q --add-port=8000/tcp && \
   firewall-cmd -q --add-port=8443/tcp && \
   firewall-cmd -q --add-port=8000/tcp --permanent && \
   firewall-cmd -q --add-port=8443/tcp --permanent"

Then, to run the container you just need to bind to those ports as -p 8000:8080 -p 8443:8443

Directories 🔗

To store some files such as the letsencrypt certificates, custom configurations or a cache with the denylists, a few directories are required:

mkdir -p ~/containers/bunkerized-nginx/{letsencrypt,cache,server-confs}

Those will be used as -v ${HOME}/containers/bunkerized-nginx/letsencrypt:/etc/letsencrypt:z -v ${HOME}/containers/bunkerized-nginx/cache:/cache:z -v ${HOME}/containers/bunkerized-nginx/server-confs:/server-confs:ro,z

Parameters 🔗

There are TONS of parameters supported by bunkerized-nginx. Some parameters can disable some features, some others enable others, etc. so grab a coffee and take a good look at the file.

In my case:

# Multisite reverse


# Nextcloud specific

podman –env-file 🔗

Reading the podman man I observed there was an --env-file parameter. So instead of having tens of -e flags, you can warp them up in a file and use just --env-file /path/to/my/envfile


systemd service 🔗

In order to run the container at boot properly, we just need to create a proper systemd file as a user such as ~/.config/systemd/user/container-bunkerized-nginx.service:

Description=Podman container-bunkerized-nginx.service

ExecStartPre=/usr/bin/rm -f /%t/%n-pid /%t/%n-cid
ExecStart=/usr/bin/podman run --conmon-pidfile /%t/%n-pid --cidfile /%t/%n-cid \
  -d --restart=always \
  -p 8000:8080 \
  -p 8443:8443 \
  -v /home/edu/containers/bunkerized-nginx/letsencrypt:/etc/letsencrypt:z \
  -v /home/edu/containers/bunkerized-nginx/cache:/cache:z \
  -v /home/edu/containers/bunkerized-nginx/server-confs:/server-confs:ro,z \
  --env-file /home/edu/containers/bunkerized-nginx/scripts/podman.env \
ExecStop=/usr/bin/podman stop -t 10 bunkerized-nginx
ExecStopPost=/usr/bin/sh -c "/usr/bin/podman rm -f `cat /%t/%n-cid`"


Notice that I didn’t use podman generate systemd because it is very specific to the container ID and I wanted more flexibility. You can read more about this in this great Running containers with Podman and shareable systemd services blog post.

Then, enable the service:

systemctl --user daemon-reload
systemctl --user enable container-bunkerized-nginx --now

This will enable the service after the first login of the user and killed after the last session of the user is closed. In order to start it after boot without requiring the user to be logged, it is required to enable lingering as:

sudo loginctl enable-linger username

Note that having the --env-file parameter makes running the container much more convinient, because it is easier to read and you can tweak the parameters in that file and just restart the service as:

systemctl --user restart container-bunkerized-nginx

Otherwise, you will need to modify the systemd unit file, run the daemon-reload command and restart the service.

Exposing it to the internet 🔗

As explained in the first post, I’m hosting all this stuff at home so I’ve configured my router, running OpenWRT, to expose only the reverse proxy ports externally (NAT) like so:

config redirect
  option dest_port '8000'
  option src 'wan'
  option name '80'
  option src_dport '80'
  option target 'DNAT'
  option dest_ip ''
  option dest 'lan'
  list proto 'tcp'

config redirect
  option dest_port '8443'
  option src 'wan'
  option src_dport '443'
  option target 'DNAT'
  option dest_ip ''
  option dest 'lan'
  list proto 'tcp'
  option name '443'

This means, that the requests incoming from the internet accessing http://my-ip will be redirected to the bunkerized-nginx container listening in port 8000, and requests accessing https://my-ip will be redirected to the bunkerized-nginx container listening in port 8443… and then, depending on the Host header, they will be redirected to the proper application container.

Next post 🔗

In the next and last post of this series, I will explain how I run the Nextcloud pod with systemd as a Kubernetes pod and how I update it.

You can read it here