systemd-docker-nginx

Run nginx inside a docker container on systemd.

Installation

1. Decide where you want to install things.

   INSTALL_DIR=/var/lib/nginx/systemd

2. Clone this repository.

   $ git clone https://github.com/agross/systemd-docker-nginx.git "$INSTALL_DIR"
   Cloning into 'systemd-docker-nginx'...

3. Install nginx systemd service unit.

   $ "$INSTALL_DIR/install"
   Linking and enabling /var/lib/nginx/systemd/nginx.service
   Created symlink /etc/systemd/system/nginx.service → /var/lib/nginx/systemd/nginx.service.
   Created symlink /etc/systemd/system/multi-user.target.wants/nginx.service → /var/lib/nginx/systemd/nginx.service.

4. Tell the nginx container about host-mounted paths for configuration, logs and HTML files.

   systemd-docker-nginx runs the nginx docker container with some required arguments, but you can also add docker run command line arguments of your own. Please see nginx.service.conf for details.

   $ systemctl edit nginx.service
   
   # Editor opens. Add these lines and save the file.
   
   [Service]
   EnvironmentFile=/path/to/your/custom/nginx.service.conf

5. Start nginx.

   $ systemctl start nginx

Optional configuration items to think about

logrotate

Depending on where your place nginx log files, you want to enable log rotatation. See NGINX_LOGS in your EnvironmentFile=.

# Get the nginx log directory via EnvironmentFile=.
$ LOGS="$(systemctl show nginx.service | \
          command grep --perl-regexp --only-matching 'EnvironmentFile=\K.*(?= \(ignore_errors=)' | \
          xargs --no-run-if-empty grep --perl-regexp --only-matching 'NGINX_LOGS=\K.*')"
$ echo $LOGS
/var/data/nginx/log

# Create logrotate script.
$ cat <<EOF > /etc/logrotate.d/nginx
$LOGS/*.log {
  daily
  dateext
  missingok
  rotate 7
  compress
  delaycompress
  notifempty
  create
  sharedscripts
  postrotate
    if systemctl is-active nginx.service > /dev/null 2>&1; then
      # Make nginx re-open its log files.
      systemctl reload nginx.service
    fi
  endscript
}
EOF

netdata

If you run netdata, it's recommended to start netdata after nginx such that netdata can detect nginx.

$ systemctl edit netdata.service

# Editor opens. Add these lines and save the file.

[Unit]
After=nginx.service

Get notified when nginx fails

You can tell systemd to start another unit whenever a unit fails. See this blog post for how it is done. Also see my systemd-unit-status-mail repository that contains everything you need.

After unit-status-mail is in place, add it to your nginx.service using a systemd override:

$ systemctl edit nginx.service

# Editor opens. Add these lines and save the file.

[Unit]
OnFailure=unit-status-mail@%n.service

Getting the whole picture

To get all nginx.service settings use

• systemctl cat nginx.service and
• systemctl show nginx.service.