Caching generally refers to the storage of web resources (HTML documents, images, and so on) in a temporary location to improve performance. Caches can be implemented by most web browsers, perimeter (intermediate) web proxies, and at the gateways of large INTERNAL networks. Transparent proxies (caches) are used by many Internet Service Providers (ISPs), reverse proxies sit in front of web servers, and the web server itself utilizes caching.

Different types of caching are explained below.

• Caching by Browsers

Web browsers implement a cache to hold recently and frequently accessed resources. It is much faster to retrieve a resource from the cache than to request it again from the web server. When deciding what to cache, browsers are usually quite well behaved and respect the caching policy dictated by the server.

There are several problems with browser caching. The size of the cache TENDS to be quite small by default and when the cache BECOMES full, the algorithm to decide what to remove is crude. Commonly, the Least Recently Used (LRU) algorithm is used to purge old items.

• Intermediate Caches:

The intermediate cache/proxy is commonly used by ISPs and larger organizations. When used by ISPs, it typically takes the form of a transparent caching proxy that silently intercepts HTTP traffic. When the client makes a request, the proxy intercepts it and checks its local cache for a copy of the resource. If none is found, it makes the request on the client’s behalf and then relays this back to the client, caching a copy itself in the process. When another client on the same ISP requests the same resource, a cached copy is returned.

Reverse Proxies Browser caches and intermediate proxies are generally for the benefit of the client, reverse proxies are usually used for the benefit of the web server. These proxies sit in front of the web server and have two purposes (although sometimes they are used for just one of these reasons): caching and load balancing. For caching, they can be used to lighten the load on the back-end web server by serving up cached versions of dynamically generated pages (thus cutting CPU USAGE). For load balancing, they can be used for load-balancing multiple back-end web servers.

Below are the different ways to control the cache.

• Controlling Caching
• Conditional GETs: Send a conditional GET request, the request knows if a local, cached copy of a resource is still valid.  The “If-Modified-Since header” is SENT in the request. If the server determines that the resource has been modified since the date given in this header, the resource is returned as normal. Otherwise, a 304 Not Modified status is returned.
• Utilizing Cache-Control and Expires Headers: Apart from “If-Modified-Since header”, there are two more headers that can control caching: Expires and Cache-Control: max-age. These headers effectively say to the client, “This resource expires on such-and-such a date. Until then, use your locally cached copy (if you have one).” The result is that the client doesn’t need to issue a conditional GET.
  • ExpiresByType image/gif "access plus 2 months“
  • ExpiresByType image/jpeg "access plus 2 months“