After the last DSP Update from 19 August 2025 (which we got two weeks later), I had quite an issue with one measure showing different / incorrect result (the logic and sub-results were fine however).
Background: SAC story optimization, where I reduced the number of used models by merging several models into one, what reduced the connections / calls from SAC to DSP and thus reduced the fetching of data by 50+%. Basically, what I did was to take the original models and merged the model-views (FACT) into one model-view and on top I created new merged model. I replaced the original models in the story with the new one, validated all measures, authorizations... Approx. 30 measures were fine, but one was off. I did a deep dive, but could not find the issue. The original model had different result as my new one, even I used the same code (copy & paste).
After like half a day of desperate attempts to find the issue on my own, double checking with colleagues, ... I have decided to build from scratch.
Long story short: at the end, I have saved-as my new model-view to V2, saved-as my new model to V2 and replaced the model-view with the V2 copy. Result was matching now.
I would not believe myself, but I did repeat the process to V3, V4, ... and did have my approach 4-eyes checked.
Solution: The fix was to create new objects as copies. I guess some obsolete / corrupt version remained in the DSP cache. So, if anyone observes issues with data and is unable to find the root cause, try to copy the objects before deep-diving.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
| User | Count |
|---|---|
| 54 | |
| 49 | |
| 22 | |
| 22 | |
| 20 | |
| 8 | |
| 7 | |
| 7 | |
| 7 | |
| 7 |