I’m new to Capella and struggle with the model validation. This is a feature that I really like and I would like to make my model pass the validation without any validation issues. I understand that I can deactivate validation rules but that seems to be cheating and I’d rather clean up my model.
For a couple of validation rules, I can simply not figure out how to clean up the model. The rules I struggle with are:
- TJ_LA_02 “[…] (Capability) should be refined towards the Logical Architecture”
I have transitioned the capabilites from the SA level to the LA level. Each capability has its corresponding Capability Realization (CR) on the LA level. The scenarios for each CR have been refined to reflect the more detailed internal structure of my system on the LA level (I’m using Exchange Scenarios here).
=> What’s the validation rule expecting me to do so that this warning goes away?
- TC_DC_11 “[…] (Logical Component) does not realize any System Component”
It is my understanding that this is the expected design workflow. When transitioning from the SA to LA level, I have to introduce new Logical Components (LCs) for the internals of my system. Capella does not allow me to add any Realized Components from the properties of the LCs, this feature is greyed out in the properties dialog.
=> How can I fix this issue so that the validation warning from rule TC_DC_11 goes away?
I’d appreciate a lot if anyone can help me with fixing these validation warnings by cleaning up my model rather than by deactivating the respective rules.
Thank you very much in advance,
Hello. Normally these validation rules should check the realization links between perspectives. I wonder if you have identified some bugs here. It will be checked.
Thank you for letting me know and checking this. Please let me know if you need more details to check if there are any bugs. Since I’m on vacation right now, my responses might be a bit slow.
Hello @JNavas, did you get any further in this? I have the same issue with the second warning @jdes describes. If you want, I can share a link to the
git repository in a private message.
Hello J. Navas,
Is it confirmed that TJ_LA_02 is supposed to check Realization links and not Refinement links?
If so, then it actually is a bug in v5.2.0 release.
The class MDCHK_Capability_Refinement_1 which is registered as the validation check class for TJ_LA_02 in its respective plugin (org.polarsys.capella.core.data.ctx.validation) does check for RefinementLink’s using the RefinementLinkExt helper class.
However, if validating with Refinement links is proper behaviour, I would like to know how to fix my models to prevent TJ_LA_02 errors as well.
With best regards,
as for TJ_LA_02, it comes from a deprecated feature, removed from Capella. Refinement capability has been removed from Capella 1.4. The rule will be removed in a next version.
as for TJ_DC_11, it will be removed too. It is useless for Logical architecture (except for the root Logical System realization).
You can disable these rules for the moment.
Hope this helps,
that’s good to know. I really like the validation feature and it will become even more useful when the built-in validation checks get more consistent and don’t generate false positives.
The following validation rules were removed: TJ_LA_02, TC_DC_11, TJ_PA_05, TJ_EPBS_02, TJ_LA_01 ( https://github.com/eclipse/capella/issues/2366)
thank you. Is the general strategy that Capella will reduce the number of built-in validation checks and that the users will have to build their own set of validation checks? Or is this commit just cleanup work and the model validation and predefined checks remain a core feature of Capella?
This is not meant as a criticism, I’d only like to understand which way you are going (and I’d personally vote for built-in validation checks as a core feature of Capella).