How to configure 'Load Balancing' Across Different Reports Servers in an Oracle Clinical / RDC Onsite 5.x Environment. (Doc ID 2301876.1)

Last updated on OCTOBER 03, 2022

Applies to:

Goal

The purpose of this note is to provide details on the 'Application Server Level' configuration required that will ensure reports jobs are submitted to the Reports Server that resides on the combined OC/RDC 5.x Application Server which the user is connected to. This brings about a kind of 'Load Balancing' where reports jobs are executed locally on the OC/RDC 5.x Application Server users are connected to.

For example :

• User 'A' > connects to OC/RDC 5.x Application Server 'A' > Reports Job will run on Server 'A'.
• User 'B' > connects to OC/RDC 5.x Application Server 'B' > Reports Job will run on Server 'B'.

Additional Information:
i.
Reports Clustering was deprecated in 10.1.2 and has been replaced with HA (High Availability). Since things are now done at the 'Application Server Level' then the local reports server on the web server has to be used. The reason for the change can be found in 'Document 343223.1 - Why Did Report Server Clustering Change in Release 10.1.2.0.2?'.


ii.
The proposed Oracle Clinical solution steps have been formulated based on the possibility of being able to associate a Reports Server with a cluster name using the REPORTS_SERVERMAP variable (in our case the %REPORT_SERVER% cluster name is a dummy). This therefore allows users to submit reports to the local reports server on the local application server. 'Document 363699.1 - Understanding Reports Execution from Forms Under High Availability' has further information on this. Basically this approach allows a System Administrator to tie a web server to a reports server in order to 'load balance' reports requests by submitting reports to the reports server that is associated with the web server the user is connected to.


iii.
The reason why the '%REPORT_SERVER%' cluster name needs to be encased in percentages is that some of the Oracle Clinical screens recognize the '%' as being an environment variable. Therefore in these cases Oracle Clinical will read the REPORT_SERVER environment variable value which was defined in the default.env file. In other cases where the % is not recognized as an environment variable then '%REPORT_SERVER%' serves the purpose of a Cluster Name which gets decoded to a reports server name using the REPORTS_SERVERMAP variable.


iv.
The 'REPORTS_SERVERMAP' entry in the rwservlets.properties file is required for RDC Onsite only.


v.
The 'REPORTS_SERVERMAP' entry in the default.env file is required for Oracle Clinical to cater for screens that do not recognize '%' as being a possible environment variable (e.g. DCI Form Generation).

Solution

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.