# Fully qualified url's in admin interface

**URL:** <https://www.ruby-forum.com/t/fully-qualified-urls-in-admin-interface/119392>\
**Category:** Radiant CMS\
**Created:** [November 6, 2007, 6:18pm UTC](https://www.ruby-forum.com/t/fully-qualified-urls-in-admin-interface/119392 "2007-11-06T18:18:40Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Erik\_Ordway](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/7bcc69/32.png) [@Erik\_Ordway](https://www.ruby-forum.com/u/Erik_Ordway)\
**Post date:** [November 6, 2007, 6:18pm UTC](https://www.ruby-forum.com/t/fully-qualified-urls-in-admin-interface/119392/1 "2007-11-06T18:18:40Z")

</div>

So I have a couple sites running Radiant .5.x and I have upgraded one  
of them to .6.3 and created another for something else. They are  
running on an W2K3 (yes I know), mongrel, and IIS with ISAPI-rewrite  
to proxy(Oh dear he really is crazy) but I have replicated this  
problem on FreeBSD and Apache.

The public parts of the sites are fine and I can get to the admin  
pages but all the links to edit/create/so user the URL base that  
mongrel is running as and giving fully qualified urls like “http://  
amago:5013/admin/page/edit/1” when the url should be “/admin/page/  
edit/1” so that the urls can be proxied.

Is this the expected behavior. It was was not under .5.x . Any  
thoughs? It seems like the " page\_new\_url(:parent\_id =\> page)" are  
using routes create the url’s and for some reason they are coming  
back a fully qualified and not relative like the links to images on  
those pages.

I think that have done something wrong but I have no idea where to  
start.

---

<div class="post-metadata">

**Author:** ![Erik\_Ordway](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/7bcc69/32.png) [@Erik\_Ordway](https://www.ruby-forum.com/u/Erik_Ordway)\
**Post date:** [November 6, 2007, 9:21pm UTC](https://www.ruby-forum.com/t/fully-qualified-urls-in-admin-interface/119392/2 "2007-11-06T21:21:26Z")

</div>

Erik,

It’s because they use \_url and not \_path. You might want to check the  
proxy headers being sent from IIS to make sure Rails is able to  
calculate the proper routes. I’m not positive which header is needed,  
but I haven’t ever had any problems with Apache or nginx.

Sean

---

<div class="post-metadata">

**Author:** ![Erik\_Ordway](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/7bcc69/32.png) [@Erik\_Ordway](https://www.ruby-forum.com/u/Erik_Ordway)\
**Post date:** [November 6, 2007, 11:06pm UTC](https://www.ruby-forum.com/t/fully-qualified-urls-in-admin-interface/119392/3 "2007-11-06T23:06:54Z")

</div>

Here was my foolishness. I forgot to put the reverse\_proxy\_fix  
plugin in place. I really wish that would get imported to rails core.

Thanks