Comparison
Data lakevsData warehouse
Data lake
everything anyone might ever need is in a bucket as files, in whatever shape it arrived, and finding the useful part is your problem.
Cheap object storage holding data in open file formats, with no enforced schema at write time. It is the answer to 'we cannot afford to decide yet': you keep everything and interpret it later, which is genuinely valuable when the questions have not been asked. What it does not give you is transactions, consistent multi-file updates or any guarantee that two readers see the same thing, which is the gap the table formats were invented to close.
Full entry →Data warehouse
the data is in typed, governed tables you query in SQL, and getting anything in means agreeing a schema first.
A managed analytical database holding modelled, typed data, optimised for large aggregate queries. You get transactions, a query planner, permissions and predictable performance in exchange for a schema up front and, historically, storage you could not read with anything else. The modern versions blur into the lakehouse, so the meaningful distinction is no longer where bytes live but whether writes are governed and typed.
Full entry →