Considering the fact that they describe two very similar types of services, it’s highly likely that most people think of data warehouses and lakes as the same thing. However, a data lake is exclusively a repository of information where everything is stored in a raw natural format. Individual objects are generally stuffed away in one as a set of abstractions that get indexed by a database. Those who opt to work with a data warehouse vendor will find that the experience is much closer to what they expect from working with a hierarchical file system.
Modern file systems like ext4 and HFSX organize documents into files that are then sorted into a series of directories. Computer software can then provide a listing of these storage locations. Since this describes the user experience provided by an overwhelming majority of commercial grade operating systems, most data warehouse vendors are going to adopt something similar. That makes their products attractive to those who plan to access files on an individual basis.
It’s also good for those who might be developing an application programming interface to communicate with everything stored in their warehouse repository. Since most APIs are designed to work with discrete file names, there’s a good chance that programmers would find writing code for this kind of environment much easier. Some of the better data warehouse vendors actually offer their own APIs to customers, which can help to further speed up information transfer rates. Users may even be able to access a remote repository as though it were one large partition they mounted locally.
