CodePudding user response:

; With table_a as (
Select "relay" as the name, 100 as fqty 
), table_b as (
Select 1 as id, 'ts - 1' as the so, "relay" as the name, 200 as fqty 
Union all select 2, 'tf - 2', 'relay, 300 
) 
Select a. *, (case when sum_curr - total<=0 then 0 when sum_curr - totalLeft the join (
Select the name, the sum (fqty) as the total 
The from table_a 
Group by name 
B) on a.n ame=b.n ame 
Cross the apply (
Select sum (fqty) as sum_curr 
The from table_b 
Where name=a.n ame and idC) 

CodePudding user response:

Have thousands of name, so there are also many

CodePudding user response:

The name with thousands of field values, how

CodePudding user response:

Create a schedule of the actual owe material C,
First-time initialization, generate related data,
Since then, the inventory list and send revised material list of operations to update a table records, C

No way is very good,
If according to the results, need skills, but also very slow, large amount of data the problem, the
As I said, it is necessary to code no loophole, carefully check each related place, but is ready to query the is very fast,