A collection in RavenDB is a set of documents with the same tag. The tag is defined in Raven-Entity-Name metadata which is filled in by the client based on the type of
an entity object that you store (the function responsible for tagging documents can by overwritten by using customizations). Note that documents that are in
the same collection can have a completely different structure, what is fine because RavenDB is schema-less.
The collection is just a virtual concept. There is no influence on how or where the documents within the same collection are stored. However the collection concept
has a great meaning for three RavenDB features: the studio, the indexes and the document key generation on the client side.
Probably the first time you encounter the collection will be accessing the studio. Then you will see that, for example, the Order entity that you have just stored is visible under
Orders collection (by default the client pluralizes the collection name based on the type name). But how does it happen that the virtual concept of the collections is
visualized in the studio. The answer is that each RavenDB database has the built-in Raven/DocumentsByEntityName index, which allows to query the database to retrieve
only documents from the specified collection. This way the studio can group the documents into the collections.
Another usage of collections is filtering documents during indexing process. When you create an index you determine what collection does it involve. During indexing process only the documents
that belong this collection are indexed, the rest is filtered out.
The default convention is that documents have the identifiers like orders/1 where orders is just a collection name and 1 is the identity value.
Also the ranges of available identity values returned by HiLo algorithm are per the collection name.