Essay: Bitmap indexes of database
Bitmap indexes are appropriate for query on non-selective columns. It is favorable when data is not frequently updated. They add a good deal of overhead to update operations and each new value of an indexed column requires a new bitmap to be creates and updates and deletes of indexed column thus require adjustment of the bitmaps since each bitmap has one value in the column. Each update locks a large number of rows, thus degrading throughput. This is its major setback (Yeh, W.H et al, 2000).
Clustering about one table is an example of table index clustering. If there exists a repetition of a portion of rows of a table with the same value, the columns can be defined as a key. This can in turn be used as a table cluster index (Girolami and Kabán, 2001). The key value obtained will be stored once encompassing all the rows sharing that key value.