为什么一个URL的页面使用另一个URL的缓存页面？

Question

Background

I'm pretty new to ASP.NET and very new to the concept of caching. I'm hoping that for someone who knows that they're doing, the problem here will be obvious.

Using MCMS 2002, I built a template.aspx page that receives a "person_id" query string and builds pages accordingly. Several different departments at the school where I work are using this template to display faculty biographies. This much is all working correctly.

The Problem

The problems start happening when I try caching the page. When a person visits person_id=16175 on one department's page, then visits the same numeric bio on a different department page, it loads the cached page instead of rebuilding it. The problem, then, is that it has all of the wrong department branding. For example:

http://health.usf.edu/medicine/obgyn/facstaf/profiles.htm?person_id=16175 http://health.usf.edu/medicine/surgery/surgery_bios.html?person_id=16175

It's the same person_id, but the URLs are obviously different. It would be great if the template would recognize the different URLs and ignore the cache. I'm assuming that the problem lies in the fact that both pages are being built using the same aspx page on the backend. Here's the OutputCache bit from the aspx page:

<%@ OutputCache Duration="86400" Location="Server" VaryByParam="person_id; section" VaryByCustom="CMSPosting" VaryByHeader="Referer" %>

I assume I'm doing something wrong here. Hopefully it will be obvious to someone who knows that they're doing. If you need more info, don't hesitate to ask. Thanks!

Answer 1

Your are most likely correct in your assumption about it being because they use the same ASPX file, I have experienced the same firsthand. To get around this issue try the following.

Try this for your output cache directive

<%@ OutputCache Location="Server" VaryByCustom="url" Duration="60" VaryByParam="*" %>

and then add this in your Global.asax file.

public override string GetVaryByCustomString(HttpContext context,string arg)
{
    if (arg == "url")
    {
         return context.Request.RawUrl;
    }
    return context.Request.RawUrl;//you can vary by other stuff here
}

A version of the page will be cached for every unique value returned by the method for a given argument. So by returning the URL you ensure the cache will not be hit for different URLs

A key thing to note is that if you do not have a single physical file resolving to multiple URLs then VaryByParam="*" would be enough to get your desired behaviour. It's because you have this URL routing that you require VaryByCustom="url"