I found the usability of "table comparison" descend after upgrade from 3.2 to 4.1 .
May someone make sure this ?
I have a ETL program , I descript its function below :
Table A is compared with table B. If some records are different with records in table B, these records would merge into table B.
Now the amount of table A and B are both 16000 , and the data are the absolute same.
I use "table comparison" to achive this function .
The "input primary key columns" are "MP_CP_ID" and "MD_BUSINESS_KEY".
In tables , all MP_CP_ID are the same, and all MD_BUSINESS_KEY are different .
In bods 4.1 :
If the order of "input primary key columns" are "md_business_key","md_cp_id", this program runs very slowly [ I haven't waited it until its ending ] .
If the order are "md_cp_id","md_business_key", it runs quickly [ around 5 seconds] .
In bods 3.2 :
No matter which one is in the first of "input primary key columns", it both runs fast .
I haven't find out explanations about this phenomeno ? Could someone tell me why ?