Hi Team, we’re using same schema in Pinot-production for 2 use-cases and one of them is external customer facing and hence critical. While the other one is internal, it faces frequent addition of attributes and history loads to schema and hence data refresh happens every month. During data refresh, data gets loaded into older partitions and if other use-case is referring those partitions, it can create issues even though use-case1 doesn’t require those new attributes or history. How can we ensure the uninterrupted data availability (without downtime) to use-case1 when data refresh happens for use-case2 ? any read replica possible ?

so you have the same table for these 2 use cases ?

yes, it’s same table for both use-cases. use-case1 has almost fixed attributes (no new addition required frequently), while other has frequent additions.

i will let others chime in ....but in my understanding you would need to have 2 different tables to solve this

please post this in #C011C9JHN7R

ok