Question

sorry if the title is not very clear. I will try to explain now:

I have two tables: table A and table B. The relation between them is one (for A table) to many (for B table). So, it’s something like master-detail situation.
I have a column ‘Amount’ in table B that is obviously decimal and a column ‘TotalAmount’ in table A.
I’m trying ti figure out how to keep the value in table A up to date. My suggestion is to make a view based on table A with aggregate query counting the Amounts from table B. Of course with the proper indexes …
But, my team-mate suggest to update the value in table A every time we change something in table B from our application.
I wonder, what will be the best solution here? May be a third variant?

Some clarification … We expected this tables to be the fastest growing tables in our database. And The table B will grow much much faster than table A. The most frequent operation in table B will be insert … and almost nothing else. The most frequent operation in table A will be select … but not only.

Answer 1

I see a number of options:

1. Use an insert trigger on table B and do the updates table A as your friend suggests. This will keep table B as up to date as possible.
2. Have a scheduled job that updates table A every x minutes (x = whatever makes sense for your application).
3. When updating table B, do an update on table A in your application logic. This may not work out if you update table B in many places.