ABSTRACT
Most databases are designed to hold data at the secondary storage medium, i.e., the disk. Consequently, the majority of the index structures are disk-based in the sense that they fetch data from the disk directly. However, with the increase in sizes of main memory storage in recent years, the concept of “in-memory” databases has become very important. In such databases, it is assumed that the entire data fits in the main memory. As a result, the index structure should be memory-based as well.
