The key idea behind push caching is to keep cached data close to those clients requesting that information. Data is dynamically mirrored as the originating server identifies where requests originate. For example, if traffic to a west coast based site started to rise because of increasing requests from the east coast, the west coast site would respond by initiating an east coast based cache.
As with adaptive caching, one main assumption of push caching is the ability to launch caches which may cross administrative boundaries. However, push caching is targeted mostly at content providers, which will most likely control the potential sites at which the caches could be deployed. Unlike adaptive caching, it does not attempt to provide a general solution for improving content access for all types of content, from all providers.