In row-oriented database systems indices are used to speed up access on columns that are often used for data selection (e.g. primary key attributes). Without indices a full table scan has to be performed when the data is stored row-wise. In column-oriented databases systems a data select operation on any attribute only requires a column scan. This is performed much faster than a complete table scan. Consequently, in many use cases additional indices are not required. The following chapter describes the usage of additional index structures in column-oriented tables, called inverted indices.
- [FSKP12]M. Faust, D. Schwalb, J. Krueger, H. Plattner, Fast lookups for in-memory column stores: group-key indices, lookup and maintenance, in ADMS ’12: Proceedings of the 3rd International Workshop on Accelerating Data Management Systems Using Modern Processor and Storage Architectures at VLDB’12, 2012Google Scholar