If you are importing very large datasets, you can divide the dataset into multiple segments, called partitions. The data in the partitions is processed simultaneously, improving the performance of your dashboards, documents, and reports.
You divide your dataset based on an attribute, called the partition attribute. When you choose a partition attribute, all the tables that contain the attribute are partitioned. It is recommended that the partition attribute be present in as many fact tables as possible.
The number of partitions that you split your dataset into should be less than or equal to the number of CPU cores on your Intelligence Server. To find the number of CPU cores on your Intelligence Server, contact your system administrator.
If you will create filters that use a search box in a the dashboard, you can improve the performance of the search by creating a search index for the attributes in your dataset. However, creating search indexes takes up additional memory in your machine.
The following steps cover partitioning and creating a search index.
This procedure assumes that you have already begun importing data, and are previewing your data. The Preview page is displayed after you select a data source to import from, configure the data source, and click Prepare Data. For a list of the data sources that you can import from, and links to steps to import data, see About importing data.
To improve performance:
On the Preview page, click All Objects View. The All Objects View dialog box is displayed, showing the attributes and metrics for your dataset, and the data tables that they are based on.
To partition the dataset:
From the Partition Attribute drop-down list, select the attribute to use to split the dataset. It is recommended that the partition attribute be present in as many fact tables as possible.
In the Number of Partitions field, type the number of partitions to split your dataset into. The number of partitions should be less than or equal to the number of CPU cores on your Intelligence Server. To find the number of CPU cores on your Intelligence Server, contact your system administrator.
To create a search index for an attribute:
In the Search Index column, enable the switch for the attribute.
Creating a search index for an attribute improves performance for filters that use a search box on that attribute. You can create search indexes on any attributes in your dataset. However, creating search indexes uses additional memory on your machine.
To save your changes, click Save. You are returned to the Preview page.
Previewing your data and specifying data import options
_____________________________
Copyright © 2019, Oracle and/or its affiliates. All rights reserved.
Legal Notices