Dashboard performance depends most on the report performance those are being used in the dashboard but there are other factors too which can affect the dashboard loading time. Below is a short list of things that should be kept in mind while developing dashboards
3. Keep the Datasets to the Appropriate Level of Aggregation. Avoid lower-level attributes if data is not required at that detailed level.
4. Test Datasets for Performance before Including Them in Dashboards. Test each dataset before including it in the dashboard to ensure it passes your performance requirements. For long running datasets, check the SQL generation and adjust the report’s VLDB settings, indexing or caching strategy to improve performance.
5. Enable Report Caching. Whenever possible, and especially with longer-running dashboards, use the caching capabilities of the Intelligence Server to save dashboard results for future use. Caches can be created/managed on a scheduled basis.
6. Use datasets that are based off Intelligent Cubes. Highly prompted reports generally do not make good caching candidates. For these reports, consider building datasets based on intelligent cubes. Not all datasets will be cube candidates (i.e. reports with conditional metrics, custom groups/ consolidations, AND NOT or OR NOT logical operators in the filter, pass through functions, report as filters).
1. Reduce the Number of Datasets and
Combine Datasets Where Possible. In general, the more datasets you include in a
dashboard, the longer it takes for that dashboard to execute, given the same
amount of data returned.
2. Remove Unused Attributes and Metrics from the Datasets.
3. Keep the Datasets to the Appropriate Level of Aggregation. Avoid lower-level attributes if data is not required at that detailed level.
4. Test Datasets for Performance before Including Them in Dashboards. Test each dataset before including it in the dashboard to ensure it passes your performance requirements. For long running datasets, check the SQL generation and adjust the report’s VLDB settings, indexing or caching strategy to improve performance.
5. Enable Report Caching. Whenever possible, and especially with longer-running dashboards, use the caching capabilities of the Intelligence Server to save dashboard results for future use. Caches can be created/managed on a scheduled basis.
6. Use datasets that are based off Intelligent Cubes. Highly prompted reports generally do not make good caching candidates. For these reports, consider building datasets based on intelligent cubes. Not all datasets will be cube candidates (i.e. reports with conditional metrics, custom groups/ consolidations, AND NOT or OR NOT logical operators in the filter, pass through functions, report as filters).
