Hi,
I want to ask for a best practise advice for such simple scenario:
We have two products table, which are related by one to many relationship:
| Products | ProductAttributeByCompany | |
| 001 | 001 | Company1 |
| 002 | 001 | Company2 |
| . | 001 | Company3 |
| . | 002 | Company1 |
| . | 002 | Company3 |
Now the "issue" with the partitioning (when using dinamic filter on "Company") is that the same row fromProductAttributeByCompany table goes only to one subscriber, however the same row fromProducts table goes to multiple subscribers, and therefor partitions cannot be used (At least it does not allow me, correct me if I'm wrong). So what is the best practise (performance wise) of creating replications:
a) Use one publication, but do not use partitioning
b) Use two publications: one for tables which cannot be partitioned, and another for tables which can be partitioned.
In my specific scenario I have around 90 tables which could be partitioned and 4 tables which can't