How Web Hosting Works on Sympl

From Sympl Wiki
Jump to navigation Jump to search

Web hosting with Sympl works somewhat automatically, creating Apache configurations and using a combination of Apache modules and redirects to serve content from the relevant directory even without a configuration, and send any unexpected traffic to the default domain.

The flow for Apache looks like this:

  1. If there is a matching configuration existing, that configuration is used.
  2. If mass hosting is not enabled, the first site found by Apache in the configuration will be served.
  3. If mass hosting is enabled, the vhost_alias module will look for an exact match for /srv/http_hostname/public/htdocs/. If it exists, it will serve the content from that directory.
  4. If there is no match, it will check for /srv/www.http_hostname/public/htdocs/, and serve content from that directory if it exists.
  5. If there is no match at all, it will serve the content from /srv/server_hostname/public/htdocs/.

Note that the mass-hosting configuration lacks some protections afforded by regular sites, such as the PHP security features in Sympl. The next time that sympl-web-configure is run (which happens hourly by default), an individual configuration for each site will be created automatically, with the PHP security features enabled.

Note that this is a little different to how Symbiosis worked, as at step 3 it would look for matches and remove elements from the start of the domain until it found a loose match, whereas Sympl only matches against example.com and www.example.com. Similarly, Symbiosis would only create a configuration for each site if it had either a specific IP set, or had SSL certificates for the domain.