Ricardo Basto wrote: > For example, the other modules try to interpret the response code of the > request before forwarding it back. My module tries the opposite: the > only case it will fake a response code different from what it got is > when the LWP request times out. Every 404, 500, 302 etc. will just be > passed back to the client. This I think is something original to my > module.
In general, the only reason people would use mod_perl for this kind of thing is so that they can modify the returned content in some way, or provide additional access control. Transparent reverse proxying is more efficiently done with mod_proxy and mod_rewrite. See this, for example: http://www.engelschall.com/pw/apache/rewriteguide/#ToC27 > And I'm neither sure if that is the best name for my module, but I think > it's target is much more people googling for "backhand" than for > "proxy". Well, mod_backhand is a load-balancer, not a proxy. It uses some proxying ability to accomplish it's goal, but it's really not about proxying. You module seems like a proxy module to me. In addition to the Apache::ReverseProxy, it looks like Apache::ProxyPass and Apache::ProxyRewrite all do essentially the same thing. Maybe you could fold your ideas into one of those. - Perrin