BusinessObjects Enterprise Administrator's Guide

Report viewing with the Cache Server and Page Server

This section describes the process for viewing a Crystal report when using the zero-client DHTML, ActiveX, or Java viewer. This process uses the Cache Server and the Page Server.

1. Upon receiving a report-viewing request, the Cache Server checks to see if it has the requested pages cached. Cached pages are stored as Encapsulated Page Format (.epf) files.
2. If a cached page for the report (.epf file) is available:
1. The Cache Server checks with the CMS to see if the user has rights to view the cached page.
2. If the user is granted the right to view the report, the Cache Server sends the cached page (.epf file) to the application server.
3. If a cached page for the report (.epf file) is unavailable:
1. The Cache Server requests new cached pages (.epf files) from the Page Server.
2. The Page Server checks with the CMS to see if the user has rights to view the report.
3. If the user is granted the right to view the report, the Page Server retrieves the report from the Input File Repository Server.
4. If the report is an instance, and the user only has View rights, the Page Server will generate pages of the report instance using the data stored in the report instance. That is, the Page Server will not retrieve the latest data from the database.

If the report is an object, the user must have View On Demand rights to view the report successfully (because the Page Server needs to retrieve data from the database).

5. If the user has sufficient rights, the Page Server generates the cached page (.epf files) and forwards them to the Cache Server.
6. The Cache Server then caches the pages (.epf files).
7. The Cache Server sends the pages (.epf files) to the application server.
4. The application server sends the report to the user's Web browser in one of two ways, depending on how the initial request was made:
• If the initial request was made through a DHTML viewer (report_view_dhtml.aspx), the viewer SDK (residing on the application server) is used to generate HTML that represents both the DHTML viewer and the report itself. The HTML pages are then returned through the web server to the user's web browser.
• If the initial request was made through an Active X or Java viewer (viewrpt.aspx), the application server forwards the cached pages (.epf files) through the web server to the report viewer software in the user's web browser.