Sequence container is a common container used to group similar tasks. It is generally used to separate processing of all dimension tables from fact tables. If we put multiple dimension tables in a sequence container, is it possible to achieve the parallel effect? An example is provided, it is OK to put the Buddha in the classroom.
1. Create a table, create table sequencetest (insert_dt datetime, insert_text varchar (50 ));
2. create a package and put three sequence container. Each sequence container puts two SQL tasks without precedence constraint for insert action to the previously created table. insert into sequencetest (insert_dt, insert_text) values (getdate (), '20140901 '). it mainly depends on the similarity of insert_dt
3. We can see that it is almost the same. However, the last task has a time difference of 0.003. Find out the specific reason, because the 0.003 seconds is exactly the SQL Server clock calculation method, measured in 0.003 units.