Classification Performance Impact

leopu

We have been using Solibri classifications extensively to structurize differing model data and are very happy with the options that are provided, however there are a few performance related issues we ran into. With growing classifications and increasing dependencies between them, the program slows down incredibly to the point where a model update takes over ten minutes.
Could you maybe give us an insight on how the classification logic works, so we can try to set them up as efficient as possible?
Some things suspected of slowing things down are:

Referencing other classifications in the input filter of a classification
Referencing other classifications as a column in the rules tab of a classification
Having too many columns in the rules tab of a classification

If we got rid of some of the classification dependencies, could that speed up the classification process?

chris

@leopu We ran into this too. And have no solution at the moment. It is very scary for us and have lot of impact on our work… Classification is a must for us!

@Solibrians FYI!

Alex.J

Same here… @Solibrians ?

manuel.furia

@leopu Everything you mentioned can slow down a classification. Also very important is what kind of data the classification operates on. For example, if the classification uses computed quantities, then those can take quite a while to compute, and will need to be recomputed every time a model is updated.