This is a draft entry for the DBMS2 analytic glossary. Please comment with any ideas you have for its improvement!
Note: Words and phrases in italics will be linked to other entries when the glossary is complete.
A data warehouse appliance is a combination of hardware and software that includes an analytic DBMS (DataBase Management System). However, some observers incorrectly apply the term “data warehouse appliance” to any analytic DBMS.
The paradigmatic vendors of data warehouse appliances are:
- Teradata, which embraced the term “data warehouse appliance” in 2008.
- Netezza — now an IBM company — which popularized the term “data warehouse appliance” in the 2000s.
Further, vendors of analytic DBMS commonly offer — directly or through partnerships — optional data warehouse appliance configurations; examples include:
- Greenplum, now part of EMC.
- Vertica, now an HP company.
- IBM DB2, under the brand “Smart Analytic System”.
- Microsoft (Parallel Data Warehouse).
Oracle Exadata is sometimes regarded as a data warehouse appliance as well, despite not being solely focused on analytic use cases.
Data warehouse appliances inherit marketing claims from the category of analytic DBMS, such as:
- Excellent performance and price/performance on complex analytic queries, …
- … often with less tuning than is needed to get acceptable analytic performance from general-purpose relational DBMS.
Further advantages specifically because they are appliances can include:
- Well-designed and -tuned hardware/software combinations.
- Easier installation or administration.
- The benefits of proprietary hardware.
Opinions differ as to whether a system has to have some of these further advantages to properly be called an “appliance”.
An alternate term for “data warehouse appliance” is analytic appliance. Reasons for using the latter term might include:
- A focus on in-memory database operations; memory-centric systems are not commonly referred to as “data warehouse appliances”.
- A desire to emphasize analytic platform features.
