The total columns will have the same datatype and size as the columns you are totalling. So if column F is smallint and the total does not fit in a smallint, you get numeric overflow.
As a last resort, you can make G and I into float - then numeric overflow is very unlikely but you could lose some significant digits. Get a realistic (and futureproof) size for the total and use this. It will then be something like:
Hi, While running the below sql, I am also getting 'Numeric Overflow Occured during computation' in Teradata. Select A ,B ,C ,D ,E ,F AS G ,H AS I from dbname.tblname WHERE D='2010-04-14' group by 1,2,3,4,5,6,7;
but I remove the 'group by..' clause and add 'distinct' after select, and can solve this problem, pls see the below sql, I do not know why.
Select distinct A ,B ,C ,D ,E ,F AS G ,H AS I from dbname.tblname WHERE D='2010-04-14';