Column

Description

Name

Name of the HPA index and its tables/indexes 

Techn. Name

For tables: technical name of the relevant index on the HPA server

Save

Status display: The system displays  to indicate that the HPA index has not yet been saved

See Creating HPA Indexes

Object Version

Status display:

 HPA index is active

 HPA index is inactive

See Activating and Filling HPA Indexes

Object Status

Status display:

 HPA index is filled

 HPA index is not filled

See Activating and Filling HPA Indexes

Number of Records

Specifies the approximate current size of the individual tables, as calculated from the database statistics

Multiple Usage

With S, X and Y tables, this indicates that one of the tables is already being used by another HPA index and therefore already exists as an index

Delta Index

Indicates that a delta index is being used for the HPA index (see Use of HPA Delta Indexes)

Last Update

Date and time of last change

Last Changed By

Name of user who made the last change

Parameters

Description

RFC Destination

RFC destination that is currently set in Customizing (transaction RSADMIN).

For more information, see Customizing under SAP Customizing Implementation Guide ® SAP NetWeaver ® Business Intelligence ® Connectivity of TREX ® Creation of RFC Destination in BI System.

You cannot change this setting here.

Block Size

This parameter is used when the HPA index is filled initially. It serves to subdivide large tables into equally large blocks. After each block, the system executes a Prepare Optimize and Commit Optimize (see Activating and Filling HPA Indexes).

Choose the corresponding entry in the context menu to display the dialog box for determining block sizes.

The larger the blocks, the quicker the overall indexing procedure.

The indexing procedure takes least time if you index all the data in a table in one step and then optimize it. For this reason, the default value for the block size is 500,000,000; larger than all the tables in an InfoCube.

However, it is useful to limit block size. If the indexing procedure terminates after a row of blocks has already been optimized (Prepare and Commit), the system can start the new job from this point. Otherwise the job for this table/index has to start again from the beginning.

Maximum Degree of Parallelization

The data is read from the database table on a package-by-package basis using OPEN CURSOR and FETCH. The system calculates package size from the width of the table and the default value in the system for package size (in bytes).

Choose the corresponding entry in the context menu to display the dialog box for determining the number of fill processes. If you set a degree of parallelization that is greater than 1, each of these packages is indexed in a new asynchronous job in dialog mode. When you start the indexing job for the package, the system reads a new package immediately. Ideally you should reduce the indexing time to the time that is required to read the data and pack it for the RFC module.

If the degree of parallelization is equal to 1, the system performs serial processing.

The optimization of the HPA index cannot be parallelized on the BI side. However, optimization automatically runs in parallel on the HPA server if the index is split.

Fact Index Parts

By default, the HPA server uses the size of the fact table on the database and an internal parameter to calculate the number of approximately equally-sized parts that the fact index is divided into.

The number of parts that is set for each fact index does not change, even if new data is loaded. You have to change the landscape of the index in order to do this. For this reason, if you know when you create the index that the size of the fact table will change significantly in a relatively short period of time (for example, increase by three of four times), we recommend that you override the internal parameters of the HPA index and split the fact index into more parts.

Choose the corresponding entry in the context menu to display the dialog box for determining fact index size.

Note: the system only consults this parameter when creating the index (see Creating HPA Indexes); changing the parameter at a later time has no effect on how data is distributed in the index.

Load Index into Main Memory

If new records are loaded to an index, the relevant index is removed from the main memory. Subsequently, the first query that requires data from this index first has to read the index from the file server, and then loads the data into the main memory. 

If you want to avoid the negative effect that this first load has on query performance, you can use this setting to determine that every time new records are loaded, the index is reloaded to the main memory immediately after the Commit Optimize.

Choose the corresponding entry in the context menu to display the Load Index into Main Memory dialog box. Input help is available.