Hi,
I have a report requirement which has premium as well as loss data. The premium base table has around 400 million records historically and the loss base table has around 5 million records. There are some filters and WHERE conditions on the tables which would limit the number of records.
I have a main SQL query for the main report with 4 sub-queries under it (i.e. 4 select statements using UNIONALL)
1. QTD calculations for Loss
2. ITD calculations for Loss
3. QTD calculations for Premium
4. ITD calculations for Premium
From a performance perspective, I wanted to know on what would be the best practice to create the dataset i.e a single master dataset would be better or having 4 separate datasets (2 for premium and 2 for loss).
I was thinking having 1 dataset would give the user the flexibility of using just 1 report and view data for premium and loss both int he same report but not sure on the performance part, hence any advice or inputs would be appreciated.
Regards,
Vishy
