Hi we've been having this problem with one of our report
The context is the following :
We have a few derived properties on failure modes :
- user_probability_class (which is computed on the failure rate of the failure mode)
- user_criticality_class (which is computed based on the severity and user_probability_class)
In the FMEA we have other derived properties on failures (~= rows in the FMEA) which simply access the user_probability_class and user_criticality_class of the represented failure modes.
Everything works great in Medini (the FMEA table correctly show everything) and in our custom FMEA report using birt when we generate it on a small FMEA.
The problems arise when we try to generate the report on big FMEAs (~= 8000 rows) :
- It takes a lot of time (which is understandable)
- Sometime it fails to compute the derived properties so they end up false
- Sometimes it even changes the derived properties in the FMEA in Medini as well (and we have to regenerate the FMEA worksheet from the model)
We noticed that closing every other projects, and freeing the maximum amount of RAM on the computer on which the report is generated sometimes makes the computation of the derived properties correct (at ~= 50% success rate).
Our questions are the following:
- Do you have any idea what we could do to remove this issue ?
- Is it possible to "freeze" the derived properties values ? (kinda like in excel where you can pause the execution of formulas)
- Is it possible to force Birt to not recompute every derived property ?
I'll provide further informations in the replies as needed.
Thanks,
Adrien Gérard