A data warehouse appliance is a combination hardware and software product that is designed specifically for analytical processing. An appliance allows the purchaser to deploy a high-performance data warehouse right out of the box.
In a traditional data warehouse implementation, the database administrator (DBA) can spend a significant amount of time tuning and putting structures around the data to get the database to perform well for large sets of users. With a data warehouse appliance, however, it is the vendor who is responsible for simplifying the physical database design layer and making sure that the software is tuned for the hardware.
When a traditional data warehouse needs to be scaled out, the administrator needs to migrate all the data to a larger, more robust server. When a data warehouse appliance needs to be scaled out, the appliance can simply be expanded by purchasing additional pug-and-play components.
A data warehouse appliance comes with its own operating system, storage, database management system (DBMS) and software. It uses massively parallel processing (MPP) and distributes data across integrated disk storage, allowing independent processors to query data in parallel with little contention and redundant components to fail gracefully without harming the entire platform. Data warehouse appliances use Open Database Connectivity (ODBC), Java Database Connectivity (JDBC), and OLE DB interfaces to integrate with other extract-transform-load (ETL) tools and business intelligence (BI) or business analytic (BA) applications.
Currently, smaller data warehouse appliance vendors seem to be concentrating on adding functionality, such as in-memory analytics, to their products in order to compete with the mega-vendors. It is anticipated, however, that all appliance vendors will be impacted by the trend toward inexpensive, high-performance, scalable virtualized data warehouse implementations that use regular hardware and open source software.