Multidimensional Data Access
All data in an InterSystems IRIS for Health™ database is stored in efficient, tree-based sparse multidimensional arrays called globals. Because they are accessed directly with no file system layer needed, InterSystems IRIS for Health globals provide very fast, flexible storage and retrieval. Globals underlie the InterSystems IRIS for Health object and SQL interfaces, and through direct access to them, InterSystems IRIS for Health supports the Key/Value and No SQL paradigms. No declarations, definitions, or allocations of storage are required for globals; they simply come into existence as data is inserted. Therefore, InterSystems IRIS for Health can easily handle dynamic data types such as XML or JSON.
The InterSystems Native APIs for Java, .NET, Python, and Node.js are lightweight interfaces that let you directly access globals so you can implement your own data structures while taking advantage of the speed and flexibility they provide.