Specifying data elements for browser-side caching

You can configure specific data elements to be cached locally in the web browser for better performance. Data elements that are cached include data that is frequently updated during common operations, such as selecting values, sorting, and filtering.

Not every dashboard feature allows data to be cached on the browser side – and some data should not be cached. However, data that is cached on the server side and the browser side work together to provide consistent results, while they reduce the server load and improving performance.

About this task

To tune your browser-side caching, you can create separate data caches, then apply different caching rules to each one. By enabling and disabling different data caches, you can discover the combination of enabled data caches that works best for your dashboard.

When browser-side caching is enabled, Cognos® Analytics uses the feature where possible. In use cases where this caching is not possible, queries are sent to the Cognos Analytics servers, but might still use server-side caching from their query services.

Note: The total size of all browser-side caches cannot exceed 100 MB of browser memory. This limit makes sure that the caches do not use a high number of browser resources. If your browser-side caches reach this size limit, any caches that did not finish generating stop generating and the dashboard displays the following error message: The size of this data cache exceeds the overall size limit for all caches (<size_in_bytes>). Apply a filter to reduce the size.

The size that is stated in the error message is the amount of browser memory that all caches consume and not the system memory of your client workstation. You can add filters to reduce the overall size of your browser-side caches and avoid reaching this size limit by clicking the Filter icon Filter icon that is associated with any categorical column. For example, if your dashboard requires data for only two years, you can filter on those two years to make sure that the cache retrieves only that data.

Limitations

Limitations of this feature include the following items:

  • Crosstab visualizations where null suppression is turned on
  • Crosstab visualizations where summary rows and columns are shown
  • Top or bottom filters
  • Navigation paths
  • Model Filters and Parameters
  • Insights/Forecasting
  • Tables with Ranks applied
  • Cognos Explorations/Stories
  • Tables with Average summaries
  • Count distinct aggregation type
  • Grouping
  • Widgets and visualizations that are set to automatically refresh

Procedure

  1. Click the Properties icon.
    Properties icon

    If you don't see the Properties icon, click the Switch to edit icon.

    Switch to edit icon
  2. For the entire dashboard or story, make sure that no other objects have focus by clicking the background, away from any visualization or object.
  3. Expand Advanced, and then next to Configure local cache, click Edit.
  4. In the Configure data caching pane, click Auto create data caches.

    One or more objects, named Data cache n appear. The number of queries, columns, and filters appears under its name.

    An image of the Configure data caching pane with a single Data cache object.

    Click the Action icon Action icon to the right of each cache to reload, delete, or duplicate the cache. Duplicating a cache can give you a starting point for a new cache that you might want to configure differently.

    Troubleshooting : If the Auto create data caches process does not complete, you might need to update certain HTTP Server configuration files. For information about resolving this issue, see 'Auto create data caches' progress starts but does not complete configuring data caches (https://www.ibm.com/support/pages/node/6997029).
  5. To see details of the columns currently designated for caching, click a Data cache n link.

    The column names are listed.

    Note: When data caches are auto-created, all available columns are initially listed for caching.
    1. To remove a column from the list, click the trash can icon Trash can icon next to the column name.
    2. To restore a column to the list, click the Add icon Add icon, select a column from the list, and then click Add,
      Tip: If nothing happens after you click the Add icon Add icon, there are no more columns available to add.
  6. To see which widgets currently have columns that are designated for caching – and therefore perform faster – click the Widgets tab.

    The widget names are listed.

    Note: When data caches are auto-created, all available widgets with cached columns are initially listed.
    1. To remove a widget from the list, click the Columns tab and remove any columns associated with the widget.
    2. To restore a widget to the list, click the Add icon Add icon, select a widget with columns not currently in this data cache, and then click Add.
      Tip: If nothing happens after you click the Add icon Add icon, there are no more widgets with uncached columns available to add.
  7. To create another data cache, in the Configure data caching pane, click the Action icon Action icon, click Add data cache, and then configure another combination of cached columns and widgets.
  8. After you determined the combination of data caches that provides the best overall performance, you can enable them and then disable the rest by choosing one or both of these options:
    • In the Configure data caching pane, click the Action icon Action icon, and then click Enable all or Disable all.
    • Select the toggle icon toggle icon next to one or more data caches to enable or disable the caching rules that you configured for that data cache.
    Tip: If you created your dashboard in Cognos Analytics 12.0, the Auto-generate on save option is enabled by default. When this option is enabled, the browser-side cache is generated each time that you save your dashboard.
    Auto-generate on save option

    If your dashboard was upgraded to version 12.0, the Auto-generate on save option is disabled by default. However, you can enable it if you want.