An Aggregating content store (AggregatingContentStore) is a content store implementation that aggregates a set of stores.
The Aggregating content store contains a primary store and a set of secondary stores. The order in which the stores appear in the list of participating stores is important. The first store in the list is known as the primary store. Content can be read from any of the stores, as if it were a single store. When the replicator goes to fetch content, the stores are searched from first to last. The stores should therefore, be arranged in order of speed.
For example, if you have a fast (and expensive) local disk, you can use this as your primary store for best performance. The old infrequently used files may be stored on lower cost, slower storage.
When replication is disabled, content is written to the primary store only. The other stores are used to retrieve content and the primary store is not updated with the content.