I was finding some strange behaviour with RFC read table approach. I know this is not suggested method for large table and involves joins.
If i am using the date format 23.06.2014, it is pulling the data from 01.01.2014. any idea ? So i had to use To_date conversion and it worked.
Individual tables are taking 1 min each to extract. if i join based on key column, it is taking around 30 mins. I know that RFC_READ_FUNCTION function module is called. So where is the join happening…at BODS or SAP. and how exactly that process works.
To mitigate point-2, I used data transfer transform with ‘Automatic’ option from one table and then joined with other table. It is giving 0 records. Strange !!! Tried with file and DB table. Same issue.
Date versus varchar will not work. You need to have a to_date() and mabye even stored in a date variable.
read table works with 1 table, not multiple. For joining SAP tables you need to use R/3 flows or an approach as you use with point 3)
Automatic is something I try to avoid. I have a strong preference for using tables for DT’s. Could be that your join does not produce results because there is no match… or because some SAP specific data such as numeric is incorrectly converted in the DT table. Try a left outer join and see what kind of results you have?
regd point 2)…can you briefly explain what happens in the background. a call will be made to RFC_READ_TABLE function and data is fetched. in case of join, it is equivalent to 2 different calls to the FM from 2 different tables. Join them in BODS and produces the output.
(BOB member since 2012-03-20)