Skip to main content

Web Caching

 
 

What is a Web Cache?

In World-Wide Web terms, a cache is a place where temporary copies of  objects are kept. Essentially, once the object pointed to by a URL has been cached, subsequent requests for the URL will result in the cached copy being returned, and little or no extra network traffic. A great deal of WWW traffic is unnecessary re-fetching from remote sites of objects which someone else at your site or in your neighbourhood already has downloaded. The local QUB cache is a large networked repository for web objects which have been previously fetched by other users. When an object is requested the local cache is searched and if the object is not cached it is downloaded from the host site. The cache will also check that the object has not been updated, and if the cached object is out of date will replace it with the newer version. Note that a web cache is often refered to as a proxy or proxy server.

Why Use a Web Cache?
Speed

Fetching a cached page should be significantly faster than fetching the page from the remote host. Popular pages with a number of images will be normally much faster to load. The downside is that an uncached object will be slightly slower to load.

Cache System Configuration
Cost

There is a potential cost saving with caching web pages as it will reduce the traffic on the external network connections. The charge for this connection may be affected by heavy or excess use.

The QUB cache is implemented on 2 SUN X4200 servers each with 2Gb of memory, these are for staff usage. The student computing centres are running qith the Qguard system which is also built on Squid but also employs URL filtering. The cache size is approximately 100Gb. The cache software is the squid package written at NLANR.