Caching techniques for use in a computer system comprising a core and at least one edge device. The core comprises at least one object addressable storage system. At least one cache is disposed logically between the core and the at least one edge device. The cache has a prefetch policy that selects from among the content units based upon at least one prefetch criterion selected from the group consisting of: a source that wrote an evaluated content unit; a size of an evaluated content unit; a content type of an evaluated content unit; when the prefetch is performed subsequent to a request to access at least one of the plurality of content units, an identity of the requestor; when the prefetch is performed subsequent to a request to access at least one of the content units stored at a first time, proximity of a time at which an evaluated content unit was stored relative to the first time; and metadata written with the evaluated content unit.