What is the best way to maximize RRC REST API performance?
It seems that the best performance is achieved by using the 'views' dataSource. This is only available via copying one of the .dta's shipped with RRC (ex. printModuleBook.dta). It is not discovered by the schema discovery tool. The query appears to grab all artifacts in the current view and then publishes them. This method is undocumented and not extensible. Conversely the examples provided by the RPE team here, http://rpeactual.com/2014/02/19/publishing-dng-rrc-modules/ have low performance (ex. 1 artifact/1+sec at first. Increasing exponentially as the number of artifacts increase), are fairly well documented and have access to most data that you would care about. Can more information be shared regarding the 'views' dataSource? Can several views be defined that can be nexsed within each other? For instance with DOORS data sources I can publish a document with a user defined table (ex. VCRM) embedded within it by using a DOORS multiple DOORS data sources configured to publish different views. |
6 answers
You can take a look at the filter by view
https://server:port/rm/publish/[artifact_format]?viewURI=id1&projectURI=id2
This works with saved views and provides a quick filter for requirements, works based on a view as the one you mention tho it is not as optimized as the printModules specific view service.
|
|
The current RRC view is used as the 'views' dataSource within RPE. How do you specify a dataSource using a saved view? This would be helpful for the last part of my question.
|
|
You can not specify the stored view on the "views" datasource from print modules. You can on any other of the public services that are documented on the wiki.
Views datasource was designed for the print only, AFAIK it is meant as an internal use datasource.
For more information check the Reportable API Query parameters
|