HPA Index 

Definition

A HPA index is a redundant data store of a BI InfoCube on the HPA server.

Use

HPA enables quick access to any data in the InfoCube with low administrative effort and is especially useful for sophisticated scenarios with unpredictable query types, high volumes of data and a high frequency of queries.

Structure

HPA Index

A HPA index contains all the data of a BI InfoCube in a compressed but not aggregated format. The HPA index stores the data at the same level of granularity as the InfoCube.

It consists of several, possibly split indexes that correspond to the tables of the enhanced star schema and a “logical” index which, depending on the star schema description, contains the metadata of the HPA index.

For more information, see the description of the HPA index in Design and Properties of HPA Indexes.

HPA Server

The HPA server is a TREX system as an installation of a HPA engine. The data of the BI InfoCube is kept and processed entirely in the main memory of the HPA server.

The HPA engine is the part of the analytics engine that manages the HPA index. This software allows the system to read data from the HPA index, add data to the HPA index, or change data. The HPA optimizer is the part of the HPA engine that ensures the best possible read access to a HPA index. For more information, see Technical Information About the HPA Engine.

Integration

Maintenance Processes for HPA Indexes

Because a HPA index, like relational aggregates, is a redundant downstream data source for improving query performance, the maintenance processes are based on and integrated with aggregate maintenance processes.

HPA Index As InfoProvider for Reporting

During query runtime, analytical engine functions such as aggregation, filtering, selection and some cell-based sorts are performed on the HPA server.