I have heard first time the term “Appliantization” from Justin Lindsey a CTO of Netezza at Netezza User Conference 2008, September, Orlando. I must admit I love this term, especially since I was involved with virtual appliance concept for geospatial. At Intergraph I was evaluating Netezza Performance Server with gaining fascinating results – truly the peroformance runs in ranges you read in Netezza marketing materials – that is 10-100 times faster than equivalent general purpose database. Gartner put Netezza into leaders sections in their magic quadrant for 2008, Netezza has quite good support for spatial types and spatial operations in their database and with UDXes you can turn the machine into domain focused Data Warehouse Appliance.
More about Netezza Spatial : http://www.netezza.com/data-warehouse-appliance-products/spatial-analytics.aspx
But let’s start from the beginning…
“One size fits all” approach doesn’t fit for high performance.
Computing Appliances are equipment with a specialized laser focus on solving targetted IT problems. In contrast to general purpose hardware and software solutions, computing appliances leverage a high level of coherence or fidelity between wired hardware and software pieces. Appliances hide the technical complexity of a system and expose the simplicity of the system. According to the Gartner definition an appliance is “a prepackaged or preconfigured balanced set of hardware, software, service and support, sold as a unit with built-in redundancy for high availability.”
Recently in the data warehouse market, new appliances have emerged with support for geospatial data, processing and present revolution (and disruptive) technology. These new appliances provide a performance boost by tackling the way large amounts of geospatial data can be effectively processed. These performance boosts are reaching orders of magnitude in comparison to general purpose database counterparts like Oracle.
Geospatially empowered Data Warehouse Appliances (DWA) with Massively Parallel Processing (MPP) architecture can scale out into the hundreds of terabytes, have capabilities to perform spatial queries in seconds instead of minutes or hours, and provide to the user new levels of experience with the affordable instant geospatial analytics.
With a huge volume of geospatially related data, there are many technical reasons to tune and assemble hardware with software and encapsulate all the complexity together into a self-contained ‘simple’ appliance with standard endpoints for interfacing. These self-contained appliances are easier to maintain and manage keeping the total cost of ownership lower than their general purpose counterparts.
2008 will be known as the year of “Appliantization.” In the data warehousing domain, appliances such as Netezza NPS, Oracle Exadata or Microsoft’s code-named project “Madison” (confluence of DataAllegro and SQL Server) are enabling technologies for high performance spatial analysis.
Simplicity is managed complexity and computing appliances just do this
∫ ∪ ∧∴ ∋ ∈ ⊇ ∧ 