Posts
Business Objects XI Server Architecture
Reference: The Complete Reference, BO XIR2 Administrator’s Guide
Actually we are used the relational database repository contained information such as universe definitions, security profiles, and corporate documents.
The use of a relational database allowed for a central source of metadata that geographically dispersed users could easily access.
Backups were easy, as DBAs could readily back up the database.
However, the relational repository also could be a bottleneck when binary files needed to be exported to or imported from the repository.
If your company had a large universe, downloading changes to that universe, while automatic for users, could take several minutes. The same is true of publishing corporate documents.
When a user published a report to Corporate Documents, the .rep file was stored as a BLOB in relational database. When another user accessed that corporate documents, the .rep file had to be extracted from the relational database and in a sense reconstructed on the system.
For large documents, this could take several minutes. .. this main drawback old versions…. This can covered advance technology with saving system.
In advance version continues to use relational repository but use the file system more extensively.
In XI, the relational repository is significantly small, relation database is used more as a method of maintaining pointers and relationships between Universe, reports, and Users. All of the universe definitions and reports are physically stored on the file system.
The best thing about this approach is the performance improvement. For Designers , exporting universe is pretty much instantaneous. For users, publishing and accessing corporate documents is as fast as saving the to disk.
In BO 6 multiple domains comes to separate folders with the XI repository and maintained via one CMS.
If the CMS is not running, then users will not be able to log in to BO.
The Input File Repository Server (FRS) handles the process of writing the results of Report servers to the repository. The Input FRS processes the request.
The Output File Repository server handles the process of serving requests for users accessing the results of a scheduled report, called an instance. When users schedule a report, a Job server will send the results of report to the Output FRS, which stores the report in a compressed format.
The Connection server provides connectivity to the data sources such as oracle, SQl server and Teradata.
A Job servers processes reports that have been scheduled to run. It will retrieve the definitions of the report from the Input FRS, connect to data source, execute the query, format the results, and save an instance of the report to the Output file Repository Server.
Job servers replace the functionality previously provided by the BCA scheduler.
There is a Web Intelligence Job server, a Desktop Intelligence Job server, and Crystal Reports Job server.
There is also a List of values Job server for Crystal Reports lists of values, not for list of values defined in Universe.
Report Servers process reports that are executed on demand or built ad hoc. There is a Web Intelligence Report Server, a Desktop Intelligence Report Server, and a Crystal Reports report server.
Cache Servers cache individual pages of reports for Desktop Intelligence and Crystal Reports users.When a User views a large report, the Cache server sends a single page to the User’s browser.
The Web Intelligence Report server has caching abilities as part of that process, so you will not see a separate Web Intelligence Cache server.
The List of values Job server processes lists of values only for Crystal Reports documents.
It does not process list of values for Web Intelligence or Desktop Intelligence documents.
Instead, the Web Intelligence report server and Desktop Intelligence Report server handle the processing for those lists of values.
0 comments:
Post a Comment