Merge Join T/r:
* This t/r is used to join 2 sources with dissimilar structures based on a joining condition.
*Difference between merge and merge join is
1) Here M-J joins sources with dissimilar structures.
2) Both t/r's should have at least one common column and that in sorted order and of same datatype
3)Here we can implement inner join , left join , right join.
Ex: Take a flat file source and table as emp , oledb source table as dept , convert the dno column into same datatype and give the o/p's of sources to sort t/r and from sort to merge join, to at last oledb destination.
Audit T/r:
* It adds some new column's like package name, task name, machine name, user name and execution time as o/p to the available i/p columns.
Ex: Try with a flat file source, link it to audit t/r.
Row Sampling and Percentage Sampling T/r:
* This t/r is used to generate specified number/percentage of random records from the source.
* It takes single i/p and gives 2 o/p's Sample records and Unsample records.
* To generate same random records for the next run use the seed value.
Ex: try with a oledb source. also try copy column t/r.
* This t/r is used to join 2 sources with dissimilar structures based on a joining condition.
*Difference between merge and merge join is
1) Here M-J joins sources with dissimilar structures.
2) Both t/r's should have at least one common column and that in sorted order and of same datatype
3)Here we can implement inner join , left join , right join.
Ex: Take a flat file source and table as emp , oledb source table as dept , convert the dno column into same datatype and give the o/p's of sources to sort t/r and from sort to merge join, to at last oledb destination.
Audit T/r:
* It adds some new column's like package name, task name, machine name, user name and execution time as o/p to the available i/p columns.
Ex: Try with a flat file source, link it to audit t/r.
Row Sampling and Percentage Sampling T/r:
* This t/r is used to generate specified number/percentage of random records from the source.
* It takes single i/p and gives 2 o/p's Sample records and Unsample records.
* To generate same random records for the next run use the seed value.
Ex: try with a oledb source. also try copy column t/r.
Comments
Post a Comment