What is the best practice for increasing the max output of JRS reports
3 answers
Hi David,
We do disk caching on the results in the 6.0 release and then the results are paged in the UI. So we should be able to accommodate large results. If you try to return all the results in the UI (i.e. not paged - 50000) you may run up against limits of the browser to handle that much data.
I can't speak to whether this is better or worse then what was done in the RRDI context.
Regards,
-Steve
50k/100k/300k records in a report sounds more of a document generation scenario. Although you could use JRS (Report Builder in 6.0) for that purpose it sounds more like a job for RPE/RRDG.
As Steve mentioned in his comment, JRS has been designed to accommodate such numbers but it is not its primary use case. In our recent test we have run reports with 50k records (out of data set with 840,000 requirements, 10,000 test cases and 60,000 test execution records) and had no problems with that. Response times for first run were between 30 seconds and minute and half (depending on exact report and test environment). Emphasizing "first run" - subsequent runs were generally faster as JRS uses various performance optimization techniques.
Now, this is obviously just an example. The numbers will depend on many things like data shape, network, actual hardware, number of concurrent users, etc. The fact that you are asking for such high numbers tells me that there is good chance such reports will be run frequently which could jeopardize performance for people running 'normal' reports. Before making such change in production environment I would do some test runs in some sort of a staging env replicating your topology and data shapes.
As Steve mentioned in his comment, JRS has been designed to accommodate such numbers but it is not its primary use case. In our recent test we have run reports with 50k records (out of data set with 840,000 requirements, 10,000 test cases and 60,000 test execution records) and had no problems with that. Response times for first run were between 30 seconds and minute and half (depending on exact report and test environment). Emphasizing "first run" - subsequent runs were generally faster as JRS uses various performance optimization techniques.
Now, this is obviously just an example. The numbers will depend on many things like data shape, network, actual hardware, number of concurrent users, etc. The fact that you are asking for such high numbers tells me that there is good chance such reports will be run frequently which could jeopardize performance for people running 'normal' reports. Before making such change in production environment I would do some test runs in some sort of a staging env replicating your topology and data shapes.
Hi David,
Yes you can change the default query results in JRS. Follow these steps:
1. Shut down JRS server if running
2. Go to the server folder JazzTeamServer\server\conf\rs\app.properties and edit the properties file
3. Add this flag to the file - query.results.limit = 50000
4. Save the changes
5. Start JRS server